JP2020141684A - Microrna biomarkers for gastric cancer diagnosis - Google Patents

Abstract

To provide alternative methods for detecting gastric cancer with less invasiveness in place of endoscopic examination.SOLUTION: Disclosed is a method for measuring the likelihood of a subject having or developing a gastric cancer or not, which comprises a step of determining the expression level of at least one miRNA having at least 90% sequence identity with one of miRNAs of specific sequences in a non-cellular bio-fluid sample obtained from the subject, where the differential expression of the miRNA as compared with a control may be indicative of the patient having gastric cancer, and the miRNA is either "up-regulated" miRNA or "down-regulated" miRNA.SELECTED DRAWING: None

Description

Cross-reference to related applications This application claims the priority benefit of Singapore Patent Application No. 10201404742W filed on August 7, 2014, the content of which is hereby incorporated by reference in its entirety for all purposes. Incorporated in the book.

The present invention relates to biochemistry, especially biomarkers. In particular, the present invention relates to cancer-related biomarkers and methods of using the biomarkers to determine the likelihood that a patient will develop cancer (eg, gastric cancer).

Gastric cancer is the fourth most common cancer (> 1 million cases per year) and is the second leading cause of cancer death worldwide. The prognosis for gastric cancer is poor, with a 5-year survival rate of less than 10%. The main reason for this is that the symptoms are usually present in the advanced state of the disease. Therefore, early detection and accurate monitoring of the disease are needed to improve clinical performance. Currently, endoscopy is the only reliable initial diagnostic method, but it is limited as a screening test due to its high cost and risk. Less invasive gastric cancer screening tests are highly desirable and are expected to eliminate unnecessary endoscopy.

Therefore, the present invention provides an alternative method for detecting gastric cancer.

Provided in one aspect is a method of detecting or diagnosing a subject's gastric cancer, or a method of determining the likelihood that a subject will develop gastric cancer, in a non-cellular biofluid sample obtained from the subject. The differential expression of the miRNA expression when compared to the control, including the step of measuring the expression level of at least one miRNA having at least 90% sequence identity with the miRNA of the above, is an indicator of the subject with gastric cancer. , The miRNA is hsa-miR-142-5p, hsa-miR-29c-3p, hsa-miR-93-5p, hsa-miR-140-5p, hsa-miR-148a-3p, hsa-miR-183 -5p, hsa-miR-29b-3p, hsa-miR-424-5p, hsa-miR-101-3p, hsa-miR-106b-3p, hsa-miR-128, hsa-miR-1280, hsa-miR -140-3p, hsa-miR-15b-3p, hsa-miR-186-5p, hsa-miR-18b-5p, hsa-miR-197-3p, hsa-miR-19a-3p, hsa-miR-19b -3p, hsa-miR-20b-5p, hsa-miR-21-3p, hsa-miR-23a-5p, hsa-miR-25-3p, hsa-miR-27a-5p, hsa-miR-29a-3p , Hsa-miR-29b-2-5p, hsa-miR-29c-5p, hsa-miR-338-5p, hsa-miR-425-3p, hsa-miR-4306, hsa-miR-450a-5p, hsa -MiR-486-5p, hsa-miR-500a-3p, hsa-miR-501-5p, hsa-miR-532-3p, hsa-miR-550a-5p, hsa-miR-579, hsa-miR-589 -5p, hsa-miR-590-5p, hsa-miR-598, hsa-miR-616-5p, hsa-miR-627, hsa-miR-629-3p, hsa-miR-629-5p, hsa-miR -93-3p, hsa-miR-195-5p, hsa-miR-18a-3p, hsa-miR-363-3p, hsa-miR-181a-2-3p, hsa-miR-16-5p, hsa-miR -501-3p, hsa-miR-23a-3p, hsa-miR-339-3p, hsa-miR-15a-5p , Hsa-miR-320b, hsa-miR-374b-5p, hsa-miR-650, hsa-miR-1290, hsa-miR-22-3p, hsa-miR-320c, hsa-miR-130a-3p, hsa -MiR-320e, hsa-miR-378a-3p, hsa-miR-9-5p, hsa-miR-200b-3p, hsa-miR-141-3p, hsa-miR-191-5p, hsa-miR-628 Is it an "up-regulated" miRNA selected from the group consisting of -5p, hsa-miR-484, hsa-miR-425-5p; or hsa-miR-103a-3p, hsa-miR-30a -5p, hsa-miR-181a-5p, hsa-miR-107, hsa-miR-26a-5p, hsa-miR-126-3p, hsa-miR-99b-5p, hsa-miR-339-5p, hsa -MiR-122-5p, hsa-miR-136-5p, hsa-miR-139-5p, hsa-miR-146a-5p, hsa-miR-154-5p, hsa-miR-193b-3p, hsa-miR -23c, hsa-miR-30b-5p, hsa-miR-337-5p, hsa-miR-382-5p, hsa-miR-409-3p, hsa-miR-411-5p, hsa-miR-485-3p , Hsa-miR-487b, hsa-miR-495, hsa-miR-885-5p, hsa-miR-99a-5p, hsa-miR-362-5p, hsa-miR-671-3p, hsa-miR-454 -3p, hsa-miR-328, hsa-miR-320a, hsa-miR-126-5p, hsa-miR-27a-3p, hsa-miR-30d-5p, hsa-miR-10a-5p, hsa-miR By any of the "down-regulated" miRNAs selected from the group consisting of -10b-5p, hsa-miR-497-5p, hsa-miR-134, and hsa-miR-150-5p. is there.

Another aspect provided is a method of detecting or diagnosing a subject's gastric cancer stage, or a method of determining the likelihood that a subject is at a stage of gastric cancer, a non-cellular biofluid obtained from said subject. At least one having at least 90% sequence identity with a microRNA (miRNA) listed in at least one of the tables selected from the group consisting of Tables 8, 15, 16, and 17 in the sample. The subject is either one of gastric cancer stage 1, stage 2, stage 3, or stage 4 due to the differential expression of miRNA expression in the sample when compared to the control, including the step of measuring the expression level of miRNA. It is a method of diagnosing that it is in.

Yet another aspect provided is a method of detecting or diagnosing gastric cancer in a subject, (a) measuring the expression level of at least one miRNA in a non-cellular biofluid sample. The miRNAs are hsa-miR-103a-3p, hsa-miR-142-5p, hsa-miR-29c-3p, hsa-miR-30a-5p, hsa-miR-107, hsa-miR-26a-5p, hsa-miR-140-5p, hsa-miR-148a-3p, hsa-miR-29b-3p, hsa-miR-126-3p, hsa-miR-181a-5p, hsa-miR-27a-5p, hsa- miR-616-5p, hsa-miR-484, hsa-miR-4306, hsa-miR-590-5p, hsa-miR-362-5p, hsa-miR-106b-3p, hsa-miR-497-5p, hsa-miR-18b-5p, hsa-miR-122-5p, hsa-miR-200b-3p, hsa-miR-197-3p, hsa-miR-486-5p, hsa-miR-99a-5p, hsa- miR-885-5p, hsa-miR-598, hsa-miR-454-3p, hsa-miR-130a-3p, hsa-miR-150-5p, hsa-miR-30d-5p, hsa-miR-10b- Consists of 5p, hsa-miR-532-3p, hsa-miR-23a-5p, hsa-miR-21-3p, hsa-miR-136-5p, hsa-miR-1280, and hsa-miR-16-5p A step having at least 90% sequence identity with the miRNA selected from the group, (b) generating a score based on the expression level of the miRNA measured in step (a), and (c) the score. In the step of predicting the likelihood that the subject has gastric cancer, the expression level of the subject is controlled by positive control (sample derived from gastric cancer subject) or negative control (sample derived from subject without gastric cancer). The score comprises the step of calculating the score by a classification algorithm to be compared with that of the subject. Do you have gastric cancer whose score falls within the positive control score; or ii. A method of identifying the likelihood of having or not having gastric cancer (no gastric cancer), where the score falls within the negative control score.

The present invention will be better understood by referring to the detailed description and considering the following non-limiting examples in combination with the accompanying drawings.

FIG. 1 shows a schematic diagram of miRNAs detected in patients with gastric cancer. FIG. 2 shows an example of a workflow diagram in which miRNAs detected in clinical samples are measured by high-throughput miRNA RT-qPCT. FIG. 2 shows the steps that occur during the process from isolation of total miRNAs from clinical samples to final data processing and statistical analysis. As described in FIG. 2, clinical samples are first subjected to an "isolation" (referring to isolating and purifying miRNA from a sample (eg, patient's serum)). Next, spike-in miRNAs (artificial synthetic miRNA mimetics (ie, small molecule single-stranded RNAs in the range of 22 to 24 bases)) are added to the sample. These spike-in miRNAs are added to the sample to monitor efficiency in various steps (including isolation, reverse transcription, enhancement, and qPCR). The miRNAs are then divided into many multiple groups. This process is divided into multiple multiplex groups (45-65 miRNAs per group) in silico by devising a "multiplication design" (miRNA assay, during the RT and augmentation processes). Designed to minimize non-specific amplification and primer-primer interactions). The sample then goes through a "multiple reverse transcription" process (a process of combining various reverse transcription primer pools and adding them to different multiple groups to synthesize cDNA). For cDNA synthesis, the PCR primer pool was combined with each cDNA pool synthesized (from some type of croup) and optimized touchdown PCR was performed to determine the total cDNA content of that group. Increase at the same time. The step of increasing the total amount of cDNA is called the "enhancement" step. Post-enhancement and prior to data processing and statistical analysis, the enhanced cDNA is a step of distributing the enhanced cDNA pool to various wells in a 384-well plate and then performing a single qPCR reaction. Refers to). Simultaneously with the measurement of the miRNA in the clinical sample, the calibration curve of the synthetic miRNA is measured together and the absolute copy number in all measurements is interpolated. Therefore, FIG. 2 provides a detailed step-by-step workflow for miRNA sample processing that occurs during the experimental section below. FIG. 3 shows a heatmap of miRNAs consistently detected in samples obtained from subjects. FIG. 3 shows a heatmap representation of all reliably detected miRNAs (listed in Table 4). MiRNA expression levels (copy / ml) were presented on a log2 scale and standardized to mean zero. The color of each point corresponds to its density. Hierarchical clustering is performed for both dimensions (miRNA and sample) based on Euclidean distance. For horizontal lines: black represents gastric cancer subjects and white represents normal / control subjects. FIG. 4 shows a heat map of controlled miRNAs. In particular, the heatmap shown in FIG. 4 shows all of the miRNAs regulated in gastric cancer, which are listed in Table 6. MiRNA expression levels (copy / ml) were presented on a log2 scale and standardized to mean zero. Grayscale corresponds to the concentration of miRNA. Hierarchical clustering is performed for both dimensions (miRNA and sample) based on Euclidean distance. For horizontal lines: black represents gastric cancer subjects and white represents normal / control subjects. FIG. 5 shows top-ranked up-regulated miRNAs (hsa-miR-142-5p) and down-regulated miRNAs (hsa-miR-99b-) in all gastric cancer subjects (regardless of subtype or stage). A boxplot showing the expression level of 5p) compared with normal / control subjects is shown. The expression level (copy / ml) is shown on the log2 scale. Box plots show the subject's miRNA expression levels, which correspond to 25%, 50%, and 75% of the expression level distribution. AUC refers to the area under the receiver operating characteristic curve. FIG. 6 shows a dot plot showing the results of correlation analysis between all reliably detected miRNAs. Based on the log2 scale expression level (copy / mL), Pearson's linear correlation coefficient was calculated among all 191 reliably detected miRNA targets (listed in Table 4). Each dot has a p-value (with respect to the difference from zero) of less than 0.01 and a correlation coefficient greater than 0.5 (left figure, positive correlation) miRNA pairs, Alternatively, it shows miRNA pairs lower than -0.5 (right figure, negatively correlated). FIG. 7 shows a dot plot showing the results of correlation analysis of gastric cancer biomarkers. Based on the log2 scale expression level (copy / mL), Pearson's linear correlation coefficient was calculated among all 75 miRNA targets that varied in gastric cancer subjects (listed in Table 6). Each dot has a p-value (when the correlation is the difference from zero) less than 0.01 and a correlation coefficient higher than 0.5 (left figure, positive correlation), or a miRNA pair. , Shows miRNA pairs lower than -0.5 (right figure, negatively correlated). Therefore, in summary, according to FIGS. 6 and 7, it was found that many cancer-related miRNAs and unrelated miRNAs have a positive correlation, which makes it difficult to select the optimum miRNA combination for gastric cancer diagnosis. are doing. FIG. 8 shows a graph in which the sensitivity and specificity of various top-rated miRNAs at various stages of gastric cancer are plotted in comparison with normal / control. More specifically, FIG. 8 shows the top up-regulated miRNAs (row 1) and down-regulated miRNAs (based on the AUC (receiver operating characteristic curve bottom area)) at various stages of the gastric cancer subject. (Line 2) shows a receiver operating characteristic curve when compared with the miRNA of a normal / control subject. FIG. 9 shows a Venn diagram illustrating the overlap between biomarkers at various stages of gastric cancer. These Venn diagrams are created based on Tables 8, 9, 10, and 11. Therefore, FIG. 9 shows that miRNAs used to determine the various stages of gastric cancer are observed in both stage 3 and stage 4 (these two stages are considered to be clinically closely related). Show that they appear to be specific to the correlating stages (except for large overlaps). FIG. 10 shows a box of diagnostic power (AUC) of a multivariate biomarker panel (3-10 miRNA types) in the discovery and validation phases during four-fold cross validation in in silico. A beard diagram is shown. Bios with 3 to 10 miRNAs using SFFS (sequence forward floating search) using a linear support vector machine as a model based on the sample discovery set and validated with another independent sample set. A marker panel was identified. Quadruple cross validation was performed multiple times. Box plots showed 25%, 50%, and 75% of the AUCs of the miRNA set for normal / control and classification of gastric cancer subjects. FIG. 10 shows the robustness of the data obtained from various multivariate biomarker panels. FIG. 11 shows a bar graph showing the average AUC of various multivariate biomarker panels in the discovery set (black bars) and validation set (rat-colored bars) during the cross-validation process. Error bars indicate the standard deviation of the AUC. To test the significance of AUC improvement in the validation set when more types of miRNAs were included in the panel, a right-hand t-test was performed to compare all adjacent murine-colored bars. *: P-value <0.05; **: p-value <0.01; ***: p-value <0.001. Therefore, FIG. 11 shows the difference in AUC between each type of miRNA in the multivariate biomarker panel set. FIG. 12 shows a schematic diagram summarizing the steps of the methods described herein, which are used to determine the likelihood that a subject has gastric cancer. FIG. 13 shows a schematic diagram summarizing the steps of the methods described herein that are used by a subject to determine the likelihood of one of the gastric cancer stages. FIG. 14 shows a schematic diagram summarizing the steps of the multivariate analysis method described herein in determining the likelihood that a subject has gastric cancer. FIG. 15 shows (A) box whiskers of controls and cancer risk scores of gastric cancer subjects; (B) application of cancer risk scores to logistic distribution for controls and gastric cancer subjects of this study. Probability distribution of gastric cancer; (C) The probability of an unknown subject (belonging to a high-risk group) with gastric cancer depends on the value of the cancer risk score (the prevalence of gastric cancer in the high-risk group is determined by the National University Hospital. (Based on National University Hospital) data, it was 0.0067); (D) The rate of increase in the probability (risk) of an unknown subject having gastric cancer at various cancer risk score levels, high risk population Compared with the prevalence of gastric cancer in Japan. The dotted line is the threshold score boundary that divides the subject into a high cancer risk group (H), a medium cancer risk group (M), or a low cancer risk group (L).

Brief Description of Tables The present invention is better understood by referring to the detailed description and considering the following non-limiting examples in combination with the attached table.

Table 1 summarizes the serum / plasma microRNA biomarker tests for gastric cancer. Table 1 summarizes the various tests that measure miRNA levels in cell-free serum / plasma samples. Only the results verified by RT-qPCR are listed in this table. "GC" refers to gastric cancer subjects and "C" refers to control subjects.

Table 2 shows the clinical information of gastric cancer subjects analyzed in the experimental section of the present disclosure. All sera were collected from 236 gastric cancer subjects prior to any treatment and stored at −80 ° C. before use. The status of H. pylori infection is shown in the column labeled "H Pylori" (referring to Helicobacter pylori).

Table 3 shows clinical information of normal subjects (controls). All normal / control subjects (237 subjects) were confirmed to be gastric cancer free by endoscopy at the time the samples were collected. Subjects were screened for endoscopy every two years and followed up for three to five years to see if they had developed gastric cancer within this period. All sera were stored at −80 ° C. before use. The status of Helicobacter pylori is shown in the column labeled "H Helicobacter pylori" (referring to Helicobacter pylori).

Table 4 shows the sequences of 191 miRNAs that are reliably detected in the serum samples of both gastric cancer subjects and normal subjects (controls). The case where miRNAs are considered to be "certainly detected" is when at least 90% of the serum samples have concentrations higher than 500 copies / ml. The miRNA was named according to the miRBase V18 release.

Table 5 summarizes the characteristics of both gastric cancer subjects and healthy controls (normal subjects / controls).

Table 6 lists the miRNAs that are differentially expressed between normal subjects (controls) and gastric cancer subjects. Comparing normal / control subjects with all gastric cancer subjects (regardless of subtype and stage), 75 miRNAs had a p-value of less than 0.01 after false discovery rate correction (Bonferroni method). It was. AUC-Area under the receiver operating characteristic curve. Magnification of change-The mean value of miRNA expression level (copy / ml) in the cancer population divided by that of the normal / control population.

Table 7 shows a comparison between this study and other literature reports. MiRNAs not listed in Table 4 (ie, miRNAs with an expression level of ≥500 copies / ml) are considered below the detection limit of this study (ND). Therefore, Table 7 shows that the majority of the differentially controlled miRNAs referred to in the art are different from the exemplary miRNAs of the present disclosure.

Table 8 lists miRNAs that are differentially expressed between normal subjects (controls) and stage 1 gastric cancer subjects. Comparing normal / control subjects with stage 1 gastric cancer subjects, the 20 miRNAs had a p-value of less than 0.01 after false discovery rate correction (Bonferroni method). AUC-Area under the receiver operating characteristic curve. Magnification of change-The mean value of miRNA expression levels in the cancer population divided by that of the normal / control population.

Table 9 lists miRNAs that are differentially expressed between normal subjects (controls) and stage 2 gastric cancer subjects. Comparing normal / control subjects with stage 2 gastric cancer subjects, 62 miRNAs had a p-value of less than 0.01 after false discovery rate correction (Bonferroni method). AUC-Area under the receiver operating characteristic curve. Magnification of change-The mean value of miRNA expression level (copy / ml) in the cancer population divided by that of the normal / control population.

Table 10 lists miRNAs that are differentially expressed between normal subjects (controls) and stage 3 gastric cancer subjects. Comparing between normal / control subjects and stage 3 gastric cancer subjects, 43 miRNAs had a p-value of less than 0.01 after false discovery rate correction (Bonferroni method). AUC-Area under the receiver operating characteristic curve. Magnification of change-The mean value of miRNA expression levels in the cancer population divided by that of the normal / control population.

Table 11 lists miRNAs that are differentially expressed between normal subjects (controls) and stage 4 gastric cancer subjects. Comparing between normal / control subjects and stage 4 gastric cancer subjects, 74 miRNAs had a p-value of less than 0.01 after false discovery rate correction (Bonferroni method). AUC-Area under the receiver operating characteristic curve. Magnification of change-The mean value of miRNA expression levels in the cancer population divided by that of the normal / control population.

Table 12 lists the miRNAs identified in the identification process of the multivariate biomarker panel. The features of miRNAs selected for biomarker panels assembled with 6, 7, 8, 9, and 10 miRNAs are summarized. The appearance rate is defined as the number of occurrences of miRNA in all panels divided by the total number of panels. Exclude panels with top 10% and bottom 10% AUC and count false-discovery biomarkers resulting from applying inaccurate data from subpopulations generated by a randomized process of cross-validation analysis. I avoided that. Among them, only miRNAs having an appearance count of 4 or more were listed. MiRNA changes at various stages of gastric cancer were defined based on Tables 6 and 8-11.

Table 13 shows statistics on the abundance of miRNAs often selected in the multivariate biomarker panel selection process. Use different numbers of miRNAs from two miRNA groups (hsa-miR-134, hsa-miR-1280, hsa-miR-27a-5p group and hsa-miR-425-5p, hsa-660-5p group) Percentage of miRNA panels to do. The top 10% and bottom 10% of the biomarker panels of 6, 7, 8, 9, and 10 miRNAs defined by the AUC of the validation set (FIG. 10) were excluded from the numbers.

Table 14 shows miRNAs (listed in Table 1) that are differentially expressed between normal / control and gastric cancer (regardless of stage) that have not been previously reported in the art. It is a list.

Table 15 is a list of miRNAs (those not previously reported in the art) that are differentially expressed between normal / control and stage 2 gastric cancer.

Table 16 is a list of miRNAs (those not previously reported in the art) that are differentially expressed between normal / control and stage 3 gastric cancer.

Table 17 is a list of miRNAs (those not previously reported in the art) that are differentially expressed between normal / control and stage 4 gastric cancer.

Table 18 shows miRNAs that are differentially expressed between normal / control and stage 1, stage 2, stage 3, stage 4, or all stages of gastric cancer (those not previously reported in the art). ) Is a list. Table 18 is a combination of Table 14, Table 8, Table 15, Table 16, and Table 17.

Table 19 shows miRNAs (ie, significant groups in Table 12) that are often selected for use in multivariate biomarker panels with altered miRNA expression levels in gastric cancer subjects (not previously reported in the art). It is a list of things).

Table 20 shows miRNAs (ie, non-significant groups in Table 12) that are often selected for use in multivariate biomarker panels where expression levels of miRNAs did not change in gastric cancer subjects (previously reported in the art). It is a list of things that have not been done).

Table 21 lists the 12 miRNAs (along with their Ki coefficient values) that are often selected in the identification process of the multivariate biomarker panel, with an incidence of> 20% and combinations thereof. Then, the cancer risk score (Equation 1) shown in FIG. 15 is calculated.

Detailed Description of the Invention MicroRNAs (miRNAs) are small molecule non-coding RNAs that act to control gene expression, and their abnormal expression is involved in the pathogenesis of various cancers (eg, gastric cancer). It is known that microRNAs can be used to determine the likelihood that a subject will develop cancer.

Therefore, one aspect provided is a method of determining the likelihood that a subject will develop or have gastric cancer. In one example, FIG. 12 is a summary of the steps of a method of determining the likelihood that a subject will develop or have gastric cancer. In one example, the method may also be for detecting or diagnosing gastric cancer in a subject. In one example, the method is an in vitro method. In one example, the methods described herein have at least 90% sequence identity with the microRNAs (miRNAs) listed in Table 18 in acellular biofluid samples obtained from subjects. It may include the step of measuring or determining the expression level of at least one miRNA, the differential expression of miRNA expression when compared to the control is an indicator that the subject has gastric cancer, and the miRNA is shown in Table 18. A method that is either a miRNA listed as "up-regulated" in Table 18 or a miRNA listed as "down-regulated" in Table 18. ..

Table 18 lists the miRNAs that can be used to determine the likelihood that a patient will develop or have gastric cancer, which are:

As used herein, the terms "miRNA", "microRNA", or "miR" are RNAs (RNA analogs) that include an endogenous non-coding gene product, and the precursor RNA transcript of that gene is small. Stem-loops can be formed where the mature "miRNA" is cleaved by the endonuclease Dicer. MiRNAs are genetically encoded, but unlike mRNAs, miRNAs regulate mRNA expression. In one example, the term "miRNA" or "microRNA" refers to a short RNA molecule in which at least 10 nucleotides and 35 or less nucleotides are covalently linked. In one example, the polynucleotide is 10 to 33 nucleotides in length, 15 to 30 nucleotides, 17 to 27 nucleotides, 18 to 26 nucleotides, ie, 10, 11, 12 , 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, or 35 nucleotides. Nucleotides, optionally, but do not contain labels or sequence extensions (eg, biotin stretches). The sequences of miRNAs described herein are provided in Table 4 provided herein, which comprises the miRNAs of SEQ ID NO: 1-SEQ ID NO: 191. It will be appreciated by those skilled in the art that miRNAs are certain types of polynucleotides having sequences containing letters such as "ATGC". It is understandable that, unless otherwise noted, the nucleotides are in the order 5'-> 3', from left to right, where "A" stands for deoxyadenoside and "C" stands for deoxycytidine. "G" stands for deoxyguanoside, and "T" stands for deoxycytidine. The letters A, C, G, and T are used to refer to the base itself, a nucleoside, or a nucleotide containing a base, which is standard in the art. The term refers to oligonucleotides consisting of naturally occurring nucleobases, sugars, and internucleotide (skeletal) bonds, as well as oligonucleotides having unnatural moieties that function similarly or specifically improve Is included. In native polynucleotides, the nucleoside bond is typically a phosphodiester bond, the component of which is referred to as the "nucleotide". The term "oligonucleotide" may also include oligonucleotides (eg, bases or sugars) that have been completely or partially modified or substituted.

It will be understood in the art of the present disclosure that, in any of the methods described herein, miRNAs do not have 100% sequence identity with the sequences of miRNAs listed in Table 4. There is. Thus, in one example, the miRNAs evaluated are listed in any one of Table 8, Table 15, Table 16, Table 17, Table 18, Table 19, or Table 20 (see Table 4). When having at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 97.5%, at least 98%, at least 99%, or at least 99.9% sequence identity with miRNA There is. In one example, the miRNA being evaluated may have one, two, three, or four nucleotide substitutions. Thus, the miRNAs used in the methods described herein are the sequences listed in Table 8, Table 15, Table 16, Table 17, Table 18, Table 19, or Table 20 (provided in Table 4). It can be detected using reagents that may be able to specifically hybridize or bind to (see sequence). The terms "hybridize" and "hybridize" as used herein are used interchangeably with the terms "specifically" and "specifically bind" and are sequenced with complementary nucleic acids. Specific non-covalent interaction, eg, interaction between a target nucleic acid sequence and a target-specific nucleic acid sequence primer or probe. In one example, the nucleic acid probe that hybridizes is one that hybridizes with 70%, 80%, greater than 90%, or 100% selectivity (ie, with one of the miRNAs described herein. Cross-hybridization may occur in less than 30%, less than 20%, less than 10%). It will be appreciated by those skilled in the art that the nucleic acid probes that "hybridize" into the miRNAs described herein can be determined given their length and composition. In one example, the nucleic acid probes that hybridize to the miRNAs described herein may not correspond to one, two, or three base pairs. In one example, the term "miRNA" described herein includes miRNAs whose 3'and / or 5'ends may be modified or digested during their production. In one example, the 3'and / or 5'ends of the miRNA pick up the modified or digested miRNA during its production in the assay of the invention. In one example, the miRNAs disclosed herein may include miRNAs that may differ from the sequences listed in Table 4 by 3, 4, or 5 nucleotides.

As used herein, the term "cancer" refers to the typical properties of cells in which cancer develops (eg, uncontrolled growth, immortality, metastatic potential, fast growth and growth rate, and certain characteristics. Refers to the presence of cells with certain morphological features) that are known in the art. In one example, the "cancer" may be gastric cancer or gastric cancer. In one example, "cancer" may include pre-cancer and malignant cancer. Thus, the term "gastric cancer" covers all stages of gastric cancer as described by the National Cancer Institute of the National Health of the National Department of Health.

In one example, what will be understood by those skilled in the art is that the methods described herein do not include steps performed by a doctor / physician. Therefore, the results obtained from the methods described herein may need to be combined with clinical data and other clinical manifestations, after which the final diagnosis made by the physician can be presented to the subject. The final diagnosis of whether a subject has gastric cancer is within the scope of the physician and is not considered part of this disclosure. Therefore, as used herein, the terms "determine," "detect," and "diagnose" are the probabilities or likelihoods that a subject will have a disease of any stage of its onset (eg, gastric cancer). Refers to the identification of, or the determination of the nature of the subject's development of the disease. In one example, "diagnosing," "determining," and "detecting" occur before the presentation of symptoms. In one example, "diagnosing," "determining," and "detecting" confirms gastric cancer in a subject whose clinician / physician is suspected of having gastric cancer (together with other clinical symptoms). Is possible.

As used herein, the term "differential expression" refers to the expression level of one or more miRNAs measured as the amount of miRNA in one sample, the same one or more miRNAs in a second sample. Refers to the difference when compared to the expression level, or when compared to a given reference value or control. Differential expression can be measured by methods known in the art (eg, array hybridization, next-generation sequencing, RT-PCR, and other methods understandable to those of skill in the art).

The group of controls used to determine if differential expression is observed may vary depending on the purpose of the methods described herein. However, as a general rule, the control group can be readily determined by one of ordinary skill in the art by any predetermined method described herein. Unless otherwise stated, in one example, the control group may be a subject with no gastric cancer-related disease. In one example, the control group may be a healthy subject with no history. In one example, the control group may be subjects known to have non-gastric cancer-related diseases (eg, but not limited to gastritis, intestinal metaplasia, intestinal atrophy, etc.). In one example, the controls may vary in sample properties and other factors (eg, gender, age, race of the subject based on the subject for whom the training dataset was obtained). In one example, the control subject may be Asian or its descendants (eg, Chinese, descendants of Chinese).

In one example, when differential expression is observed, the miRNAs listed as "upregulated" in Table 18 are obtained from subjects suspected of having gastric cancer when compared to controls. This is the case when it is found to be up-regulated in the sample. Thus, in one example, the miRNA upregulation listed as "upregulated" in Table 18 as compared to the control may be an indicator that the subject has gastric cancer, or it. Diagnose the subject as having gastric cancer.

In any example or method described herein, the term "upregulated" as used herein refers to an increase in miRNA expression levels, or more copies, as compared to controls. Refers to the detection of miRNA. In one example, miRNAs can be considered "upregulated" when the rate of change of expression level relative to the control is at least about 1.01 or greater. In one example, the p-value in the statistical test for that change may be less than 0.05. In one example, the p-value can be calculated using a statistical test known in the art (t-test (p-value <0.01)). In another example, the statistical test is a t-test (p-value <0.01), corrected for false discovery rate (FDR) using the Bonferroni-type multiple comparison method known in the art. is there. This approach was used exemplary in the examples below. In one example, miRNAs can be considered "upregulated" when the rate of change of expression level relative to the control is at least about 1.02 to about 2.5. In one example, as illustrated in Table 6, miRNAs are considered to be "upregulated" when the rate of change in expression level relative to the control is at least about 1.05-about 1.53. Can be done. In one example, miRNAs have at least about 1.10-fold, about 1.15-fold, about 1.20-fold, about 1.25-fold, about 1.30-fold, and about 1. When it is 35 times, about 1.40 times, about 1.45 times, and about 1.50 times, it can be considered to be "up-regulated". The controls used herein may be subjects without gastric cancer.

In one example, when differential expression is observed, the miRNAs listed as "down-regulated" in Table 18 are obtained from subjects suspected of having gastric cancer when compared to controls. This is the case when it is found that the sample is down-regulated. Thus, in one example, the down-regulation of miRNA listed as "down-regulated" in Table 18 as compared to the control may be an indicator that the subject has gastric cancer, or it. It is also possible to diagnose that the subject has gastric cancer.

In any example or method described herein, the term "downregulated" as used herein refers to a reduction in miRNA expression levels or a smaller copy of miRNA expression levels compared to controls. Refers to the detection of miRNA. In one example, miRNAs can be considered "down-regulated" if the rate of change in expression level relative to the control is shown to be at least about 0.99 or less. In one example, the p-value in the statistical test for that change may be less than 0.05. In one example, the p-value can be calculated using a statistical test known in the art (t-test (p-value <0.01)). In another example, the statistical test is a t-test (p-value <0.01), corrected for false discovery rate (FDR) using the Bonferroni-type multiple comparison method known in the art. is there. This approach was used exemplary in the examples below. In one example, miRNAs can be considered "down-regulated" when the rate of change of expression level relative to the control is at least about 0.98 to about 0.3. In one example, as illustrated in Table 6, miRNAs are "down-regulated" when the rate of change in expression level relative to the control is shown to be at least about 0.5 to about 0.92. Can be thought of. In one example, miRNA expression has at least about 0.95, about 0.90, about 0.85, about 0.80, about 0.75, and about 0, with respect to control. When it is .70 times, about 0.65 times, about 0.60 times, and about 0.55 times, it can be considered to be "down-regulated". The controls used herein may be subjects without gastric cancer.

The term "about" is used herein to determine the inherent variability or test of errors known in the art (those that may result from the use of equipment or methods employed to determine the value). It is shown to include the variability that exists between the subjects. Thus, in one example, the term "about" is ± 5%, ± 4%, ± 3 of the values referred to in the context of the rate of change of expression as used throughout the disclosure described herein. It means%, ± 2%, ± 1%, or ± 0.5%.

In one example, the methods described herein can measure the differential expression of one or more miRNAs listed in Table 18. Thus, in one example, the method comprises at least two, at least three, at least four, at least five, at least six, at least seven, at least eight, and at least nine miRNAs listed in Table 18. , At least 10 types, at least 11 types, at least 12 types, at least 13 types, at least 14 types, at least 15 types, at least 20 types, at least 25 types, at least 30 types, at least 35 types, at least 40 types, at least 45 types, at least 50 types, at least 55 types, at least 60 types, at least 65 types, at least 70 types, at least 75 types, at least 80 types, at least 85 types, at least 90 types, at least 95 types, at least 100 types, at least 105 types, at least 107 types , Or all differential expression can be measured. In one example, the method comprises 1 to 108 miRNAs listed in Table 18 or 2 to 10, 11 to 30, 31 to 50, 50 to 70, 71 to 90, listed in Table 18. Alternatively, 91 to 108 types of miRNA can be measured. In one example, the method is capable of measuring the differential expression of all miRNAs listed in Table 18.

In one example, the methods described herein are listed as "up-regulated" in Table 18 and at least one miRNA listed as "down-regulated" in Table 18. It is possible to measure the differential expression of at least one miRNA that is produced. In one example, the method comprises at least two, at least three, at least four, at least five, at least six, and at least seven miRNAs listed as "up-regulated" in Table 18. , At least 8 types, at least 9 types, at least 10 types, at least 11 types, at least 12 types, at least 13 types, at least 14 types, at least 15 types, at least 20 types, at least 25 types, at least 30 types, at least 35 types, at least 40, at least 45, at least 50, at least 55, at least 60, at least 65, or 70 differential expressions can be measured and are "down-regulated" in Table 18. At least 2 types, at least 3 types, at least 4 types, at least 5 types, at least 6 types, at least 7 types, at least 8 types, at least 9 types, at least 10 types, at least 11 types, and at least 12 types of miRNAs listed as , At least 13 types, at least 14 types, at least 15 types, at least 20, at least 25 types, at least 30, at least 35 types, or 38 types of differential expression can be measured.

In one example, the methods described herein include measuring the expression levels of microRNAs (miRNAs) listed in Table 14 in acellular bodily fluid samples obtained from subjects. There is.

Table 14 lists the miRNAs that can be used to determine the likelihood that a patient will develop or have gastric cancer, which are:

As is known in the art, when proliferative disease progresses (with or without treatment), the clinical manifestations of the disease vary between stages of progression. Therefore, to assist those skilled in the art in determining how gastric cancer / gastric cancer has progressed, the National Cancer Institute of the National Department of Health has provided general clinical manifestations of the stages observed in gastric cancer / gastric cancer. Published. The stages of gastric cancer include, but are not strictly limited to, the following stages as described below.

Stage I: In stage I, the cancer has already formed on the inner wall of the mucosa (innermost layer) of the gastric wall. Stage I is divided into stage IA and stage IB, depending on where the cancer has spread.

Stage IA: Cancer may have already spread into the submucosa of the gastric wall (the panniculus adjacent to the mucosa).

Stage IB: The cancer may have already spread into the submucosa of the gastric wall (the tissue layer adjacent to the mucosa) and is found in one or two lymph nodes near the tumor; or the cancer. Has already spread to the muscle layer of the gastric wall.

Stage II: Stage II gastric cancer is divided into stage IIA and stage IIB, depending on where the cancer has spread.

Stage IIA: Has the cancer already spread to the subserosal layer of the gastric wall (the tissue layer adjacent to the serosa); has it already spread to the muscle layer of the gastric wall and is found in one or two lymph nodes near the tumor? Alternatively, it has already spread to the submucosa of the gastric wall (the tissue layer adjacent to the mucosa) and is found in 3 to 6 lymph nodes near the tumor.

Stage IIB: Has the cancer already spread to the serosa of the gastric wall (the outermost layer); has already spread to the submucosa of the gastric wall (the tissue layer adjacent to the serosa) and has one or two lymphs near the tumor? Is it found in the nodes; has already spread to the muscular layer of the gastric wall and has been found in 3 to 6 lymph nodes near the tumor; or has already spread to the submucosa of the gastric wall (the tissue layer adjacent to the mucosa) It is found in more than 7 lymph nodes near the tumor.

Stage III: Stage III gastric cancer is divided into stage IIIA, stage IIIB, and stage IIIC, depending on where the cancer has spread.

Stage IIIA: Is the cancer already spread to the serosal (outermost) layer of the gastric wall and found in one or two lymph nodes near the tumor; the subserosal layer of the gastric wall (tissue layer adjacent to the serosa)? It has already spread to 3-6 lymph nodes near the tumor; or it has already spread to the muscular layer of the gastric wall and is found in 7 or more lymph nodes near the tumor.

Stage IIIB: Cancer has already spread to nearby organs (eg, spleen, transverse colon, liver, serosa, pancreas, kidney, adrenal gland, or small intestine) and has been found in one or two lymph nodes near the tumor. Is it already spread to the serosa of the stomach wall (outermost layer) and found in 3 to 6 lymph nodes near the tumor; or already in the subserosal layer of the stomach wall (tissue layer adjacent to the serosa) It is widespread and is found in more than 7 lymph nodes near the tumor.

Stage IIIC: Cancer has already spread to nearby organs (eg, spleen, transverse colon, liver, serosa, pancreas, kidney, adrenal gland, or small intestine) and is found in three or more lymph nodes near the tumor. In some cases; or it has already spread to the serosa of the stomach wall (the outermost layer) and is found in more than 7 lymph nodes near the tumor.

Stage IV: In stage IV, the cancer has already spread to the distal part of the body.

As will be appreciated by those skilled in the art, deciding whether a subject is at a particular stage provides the clinician with clear guidelines on how to best address any treatment or palliative care. May be useful. Therefore, in one aspect, the present disclosure provides a method of detecting or diagnosing a gastric cancer stage in a subject. In one example, the method determines the likelihood that a subject is at a stage of gastric cancer. In this regard, the method is selected from the group consisting of Table 8, Table 9, Table 15, Table 10, Table 16, Table 11, or Table 17 in the acellular biofluid sample obtained from the subject. Including the step of measuring the expression level of a miRNA having at least 90% sequence identity with a microRNA (miRNA) listed in at least one of the above, miRNA expression in the sample obtained from the subject when compared to the control. The differential expression of may be an indicator of the likelihood that the subject is in any one of stage 1, stage 2, stage 3, or stage 4 of gastric cancer, or due to said differential expression. The subject may be diagnosed with gastric cancer at any one of stage 1, stage 2, stage 3, or stage 4. The controls used herein may be subjects without gastric cancer. In one example, FIG. 13 is a summary of the steps of how a subject determines the likelihood of developing or having one of the stages of gastric cancer.

In one example, when differential expression is observed, the subject is in gastric cancer stage 1 with a miRNA listed as "upregulated" in Table 8 when compared to the control. The miRNA can be used as an indicator of gastric cancer, or the miRNA can be used to diagnose that the subject is in stage 1 gastric cancer, and the miRNA is listed as "up-regulated" in Table 8. If at least one of the above can be included. Table 8 is described in more detail in the experimental section. In one example, miRNAs can be considered "upregulated" when the rate of change of expression level relative to the control is at least about 1.01 to about 2.5 or greater. In one example, the p-value in the statistical test for that change may be less than 0.05. In one example, the p-value can be calculated using a statistical test known in the art (t-test (p-value <0.01)). In another example, the statistical test is a t-test (p-value <0.01), corrected for false discovery rate (FDR) using the Bonferroni-type multiple comparison method known in the art. is there. This approach was used exemplary in the examples below. In one example, miRNAs have at least about 1.10-fold, about 1.15-fold, about 1.20-fold, about 1.25-fold, about 1.30-fold, and about 1. When it is 35 times, about 1.40 times, about 1.45 times, and about 1.50 times, it can be considered to be "up-regulated". In one example, as illustrated in Table 8, miRNAs are considered to be "upregulated" when the rate of change in expression level relative to the control is at least about 1.11 to about 1.57. Can be done.

In one example, the methods described herein are capable of measuring the differential expression of one or more miRNAs listed as "upregulated" in Table 8. Thus, in one example, the method is listed as "up-regulated" in Table 8 at least 2, at least 3, at least 4, at least 5, at least 6, and at least 7. , Or the differential expression of eight miRNAs can be measured.

In one example, when differential expression is observed, the subject is in gastric cancer stage 1 with a miRNA listed as "down-regulated" in Table 8 when compared to the control. It is possible that the subject can be diagnosed as having gastric cancer stage 1 by the miRNA. In one example, the miRNA can include at least one of the miRNAs listed as "down-regulated" in Table 8. In one example, miRNAs can be considered "down-regulated" when the rate of change of expression level relative to the control is at least about 0.9 or less. In one example, the p-value in the statistical test for that change may be less than 0.05. In one example, the p-value can be calculated using a statistical test known in the art (t-test (p-value <0.01)). In another example, the statistical test is a t-test (p-value <0.01), corrected for false discovery rate (FDR) using the Bonferroni-type multiple comparison method known in the art. is there. This approach was used exemplary in the examples below. In one example, miRNA expression has at least about 0.95, about 0.90, about 0.85, about 0.80, about 0.75, and about 0, with respect to control. When it is .70 times, about 0.65 times, about 0.60 times, and about 0.55 times, it can be considered to be "down-regulated". In one example, as illustrated in Table 8, miRNAs are considered to be "down-regulated" when the rate of change in expression level relative to the control is at least about 0.68 to about 0.84. Can be done.

In one example, the methods described herein are capable of measuring the differential expression of one or more miRNAs listed as "down-regulated" in Table 8. Thus, in one example, the method is listed as "down-regulated" in Table 8, at least 2, at least 3, at least 4, at least 5, at least 6, and at least 7. , At least 8, at least 9, at least 10, at least 11 or 12 miRNAs can be measured for differential expression.

In one example, the methods described herein are listed as "up-regulated" in Table 8 and at least one miRNA listed as "down-regulated" in Table 8. It is possible to measure the differential expression of at least one miRNA that is produced. Therefore, the method comprises at least two, at least three, at least four, at least five, at least six, at least seven, or eight types listed as "up-regulated" in Table 8. Differential expression of miRNAs can be measured and at least two, at least three, at least four, at least five, at least five of the miRNAs listed as "down-regulated" in Table 8. Differential expression of 6, at least 7, at least 8, at least 9, at least 10, at least 11, or 12 can be measured.

In one example, when differential expression is observed and when compared to the control, the miRNA listed as "upregulated" in Table 9 or Table 15 is the subject of gastric cancer stage 2 Or if the miRNA can be used to diagnose that the subject is in stage 2 gastric cancer. Table 9 is described in more detail in the Experimental section. In one example, the miRNA can include at least one of the miRNAs listed as "up-regulated" in Table 15.

Table 15 lists the miRNAs used to determine the likelihood that a patient will develop (affected) or have stage 2 gastric cancer, which are:

In one example, miRNAs can be considered "upregulated" when the rate of change of expression level relative to the control is at least about 1.01 to about 2.5 or greater. In one example, the p-value in the statistical test for that change may be less than 0.05. In one example, the p-value can be calculated using a statistical test known in the art (t-test (p-value <0.01)). In another example, the statistical test is a t-test (p-value <0.01), corrected for false discovery rate (FDR) using the Bonferroni-type multiple comparison method known in the art. is there. This approach was used exemplary in the examples below. In one example, miRNAs have at least about 1.10-fold, about 1.15-fold, about 1.20-fold, about 1.25-fold, about 1.30-fold, and about 1. When it is 35 times, about 1.40 times, about 1.45 times, and about 1.50 times, it can be considered to be "up-regulated". In one example, miRNAs are considered to be "upregulated" when the rate of change in expression level relative to the control is at least about 1.13 to about 2.13, as illustrated in Table 9. Can be done.

In one example, the methods described herein are capable of measuring the differential expression of one or more miRNAs listed as "upregulated" in Table 9 or Table 15. Thus, in one example, the method is listed as "up-regulated" in Table 9 or 15, at least 2, at least 3, at least 4, at least 5, at least 6, and at least. 7 types, at least 8 types, at least 9 types, at least 10 types, at least 11 types, at least 12 types, at least 13 types, at least 14 types, at least 15 types, at least 20 types, at least 25 types, at least 30 types, at least 35 types , At least 40, at least 45, at least 50, or 54 miRNAs can be measured for differential expression.

In one example, when differential expression is observed and when compared to controls, the miRNA listed as "down-regulated" in Table 9 or Table 15 is the subject of gastric cancer stage 2 Or if the miRNA can be used to diagnose that the subject is in stage 2 gastric cancer. In one example, the miRNA can include at least one of the miRNAs listed as "down-regulated" in Table 15. In one example, miRNAs can be considered "down-regulated" when the rate of change of expression level relative to the control is at least about 0.9 or less. In one example, the p-value in the statistical test for that change may be less than 0.05. In one example, the p-value can be calculated using a statistical test known in the art (t-test (p-value <0.01)). In another example, the statistical test is a t-test (p-value <0.01), corrected for false discovery rate (FDR) using the Bonferroni-type multiple comparison method known in the art. is there. This approach was used exemplary in the examples below. In one example, miRNA expression has at least about 0.95, about 0.90, about 0.85, about 0.80, about 0.75, and about 0, with respect to control. When it is .70 times, about 0.65 times, about 0.60 times, and about 0.55 times, it can be considered to be "down-regulated". In one example, miRNAs are considered to be "down-regulated" when the rate of change in expression level relative to the control is at least about 0.56 to about 0.87, as illustrated in Table 9. Can be done.

In one example, the methods described herein are capable of measuring the differential expression of one or more miRNAs listed as "down-regulated" in Table 9 or Table 15. Thus, in one example, the method is listed as "down-regulated" in Table 9 or Table 15, at least two, at least three, at least four, at least five, at least six. Differential expression of at least 7 or 8 miRNAs can be measured.

In one example, the methods described herein are at least one miRNA listed as "up-regulated" in Table 9 or Table 15 and "down-regulated" in Table 9 or Table 15. It is possible to measure the differential expression of at least one miRNA listed as ". In one example, the method comprises at least two, at least three, at least four, at least five, at least six types of miRNA listed as "upregulated" in Table 9 or Table 15. At least 7 types, at least 8 types, at least 9 types, at least 10 types, at least 11 types, at least 12 types, at least 13 types, at least 14 types, at least 15 types, at least 20 types, at least 25 types, at least 30 types, at least 35 types Differential expression of at least 40, at least 45, at least 50, or 54 species can be measured; and miRNAs listed as "down-regulated" in Table 9 or Table 15. It is possible to measure the differential expression of at least 2 types, at least 3 types, at least 4 types, at least 5 types, at least 6 types, at least 7 types, or 8 types.

In one example, if the method described herein can provide a diagnosis or indicator as to whether a subject may have stage 2 gastric cancer, the method is limited. However, expression of any one or more miRNAs, including miR-20a-5p, miR-223-3p, miR-17-5p, miR-106b-5p, miR-423-5p, and miR-21-5p. An additional step of comparing levels may be included.

In one example, when differential expression is observed and when compared to controls, miRNAs listed as "upregulated" in Table 10 or Table 16 are used by the subject in gastric cancer stage 3 Or if the miRNA can be used to diagnose that the subject is in stage 3 gastric cancer. Table 10 is described in more detail in the experimental section. In one example, the miRNA can include at least one of the miRNAs listed as "up-regulated" in Table 16.

Table 16 lists the miRNAs used to determine the likelihood that a patient will develop (affected) or have stage 3 gastric cancer, which are:

In one example, miRNAs can be considered "upregulated" when the rate of change of expression level relative to the control is at least about 1.01 to about 2.5 or greater. In one example, the p-value in the statistical test for that change may be less than 0.05. In one example, the p-value can be calculated using a statistical test known in the art (t-test (p-value <0.01)). In another example, the statistical test is a t-test (p-value <0.01), corrected for false discovery rate (FDR) using the Bonferroni-type multiple comparison method known in the art. is there. This approach was used exemplary in the examples below. In one example, miRNAs have at least about 1.10-fold, about 1.15-fold, about 1.20-fold, about 1.25-fold, about 1.30-fold, and about 1. When it is 35 times, about 1.40 times, about 1.45 times, and about 1.50 times, it can be considered to be "up-regulated". In one example, miRNAs are considered to be "upregulated" when the rate of change in expression level relative to the control is at least about 1.20 to about 1.93, as illustrated in Table 10. Can be done.

In one example, the methods described herein are capable of measuring the differential expression of one or more miRNAs listed as "upregulated" in Table 10 or Table 16. Thus, in one example, the method is listed as "up-regulated" in Table 10 or Table 16, at least two, at least three, at least four, at least five, at least six. At least 7 types, at least 8 types, at least 9 types, at least 10 types, at least 11 types, at least 12 types, at least 13 types, at least 14 types, at least 15 types, at least 20 types, at least 25 types, at least 30 types, at least 35 types It is possible to measure the differential expression of 37 or 37 miRNAs.

In one example, if differential expression is observed, the miRNA listed as "down-regulated" in Table 10 or Table 16 when compared to the control will cause the subject to stage 3 gastric cancer. It may be an indicator of something, or it is possible to diagnose that the subject is in stage 3 gastric cancer by the miRNA. In one example, the miRNA can include at least one of the miRNAs listed as "down-regulated" in Table 16. In one example, miRNAs can be considered "down-regulated" when the rate of change of expression level relative to the control is at least about 0.9 or less. In one example, the p-value in the statistical test for that change may be less than 0.05. In one example, the p-value can be calculated using a statistical test known in the art (t-test (p-value <0.01)). In another example, the statistical test is a t-test (p-value <0.01), corrected for false discovery rate (FDR) using the Bonferroni-type multiple comparison method known in the art. is there. This approach was used exemplary in the examples below. In one example, miRNA expression has at least about 0.95, about 0.90, about 0.85, about 0.80, about 0.75, and about 0, with respect to control. When it is .70 times, about 0.65 times, about 0.60 times, and about 0.55 times, it can be considered to be "down-regulated". In one example, as illustrated in Table 10, miRNAs are considered to be "down-regulated" when the rate of change in expression level relative to the control is at least about 0.72-about 0.88. Can be done.

In one example, the methods described herein are capable of measuring the differential expression of one or more miRNAs listed as "down-regulated" in Table 10 or Table 16. Thus, in one example, the method comprises at least two, at least three, at least four, at least five, or six types listed as "down-regulated" in Table 10 or Table 16. The differential expression of miRNA can be measured.

In one example, the methods described herein are at least one miRNA listed as "up-regulated" in Table 10 or 16 and "down-regulated" in Table 10 or 16. It is possible to measure the differential expression of at least one miRNA listed as ". In one example, the method is listed as "up-regulated" in Table 10 or 16, at least 2, at least 3, at least 4, at least 5, at least 6, and at least 7. , At least 8 types, at least 9 types, at least 10 types, at least 11 types, at least 12 types, at least 13 types, at least 14 types, at least 15 types, at least 20 types, at least 25 types, at least 30 types, at least 35 types, or The differential expression of 37 miRNAs can be measured and at least two, at least three, at least four, listed as "down-regulated" in Table 10 or Table 16. Differential expression of at least 5 or 6 miRNAs can be measured.

In one example, Table 16 is used for the methods described herein that provide a diagnosis or indicator of whether a subject may have stage 3 gastric cancer. The method may further include, but is not limited to, the step of comparing the expression levels of any one or more miRNAs, including, but not limited to, miR-21-5pm, miR-223-3p, and miR-423-5p.

In one example, when differential expression is observed and when compared to the control, the miRNA listed as "upregulated" in Table 11 or Table 17 is the subject of gastric cancer stage 4 It may be an indicator that the subject is in stage 4, or the miRNA can be used to diagnose that the subject is in stage 4 gastric cancer. Table 10 is described in more detail in the experimental section. In one example, the miRNA can include at least one of the miRNAs listed as "upregulated" in Table 17.

Table 17 lists the miRNAs used to determine the likelihood that a patient will develop (affected) stage 4 gastric cancer, which are:

In one example, miRNAs can be considered "upregulated" when the rate of change of expression level relative to the control is at least about 1.01 to about 2.5 or greater. In one example, the p-value in the statistical test for that change may be less than 0.05. In one example, the p-value can be calculated using a statistical test known in the art (t-test (p-value <0.01)). In another example, the statistical test is a t-test (p-value <0.01), corrected for false discovery rate (FDR) using the Bonferroni-type multiple comparison method known in the art. is there. This approach was used exemplary in the examples below. In one example, miRNAs have at least about 1.10-fold, about 1.15-fold, about 1.20-fold, about 1.25-fold, about 1.30-fold, and about 1. When it is 35 times, about 1.40 times, about 1.45 times, and about 1.50 times, it can be considered to be "up-regulated". In one example, miRNAs are considered to be "upregulated" when the rate of change in expression level relative to the control is at least about 1.12 to about 1.86, as illustrated in Table 11. Can be done.

In one example, the methods described herein are capable of measuring the differential expression of one or more miRNAs listed as "upregulated" in Table 11 or Table 17. Thus, in one example, the method is listed as "up-regulated" in Table 11 or 15, at least two, at least three, at least four, at least five, at least six, at least. 7 types, at least 8 types, at least 9 types, at least 10 types, at least 11 types, at least 12 types, at least 13 types, at least 14 types, at least 17 types, at least 20 types, at least 25 types, at least 30 types, at least 35 types , At least 40, at least 45, or 46 miRNAs can be measured for differential expression.

In one example, when differential expression is observed and when compared to the control, the miRNA listed as "down-regulated" in Table 11 or Table 17 is the subject of gastric cancer stage 4 It may be an indicator that the subject is in stage 4, or the miRNA can be used to diagnose that the subject is in stage 4 gastric cancer. In one example, the miRNA can include at least one of the miRNAs listed as "down-regulated" in Table 17. In one example, miRNAs can be considered "down-regulated" when the rate of change of expression level relative to the control is at least about 0.9 or less. In one example, the p-value in the statistical test for that change may be less than 0.05. In one example, the p-value can be calculated using a statistical test known in the art (t-test (p-value <0.01)). In another example, the statistical test is a t-test (p-value <0.01), corrected for false discovery rate (FDR) using the Bonferroni-type multiple comparison method known in the art. is there. This approach was used exemplary in the examples below. In one example, miRNA expression has at least about 0.95, about 0.90, about 0.85, about 0.80, about 0.75, and about 0, with respect to control. When it is .70 times, about 0.65 times, about 0.60 times, and about 0.55 times, it can be considered to be "down-regulated". In one example, as illustrated in Table 11, miRNAs are considered to be "down-regulated" when the rate of change in expression level relative to the control is at least about 0.48 to about 0.90. Can be done.

In one example, the methods described herein are capable of measuring the differential expression of one or more miRNAs listed as "down-regulated" in Table 11 or Table 17. Thus, in one example, the method is listed as "down-regulated" in Table 11 or 17, at least 2, at least 3, at least 4, at least 5, at least 6, and at least. Differences between 7 types, at least 8 types, at least 9 types, at least 10 types, at least 11 types, at least 12 types, at least 13 types, at least 14 types, at least 15 types, at least 20, at least 25 types, or 28 types of miRNA The next expression can be measured.

In one example, the methods described herein are at least one miRNA listed as "up-regulated" in Table 11 or 17 and "down-regulated" in Table 11 or 17. It is possible to measure the differential expression of at least one miRNA listed as ". In one example, the method is listed as "up-regulated" in Table 11 or 17, at least 2, at least 3, at least 4, at least 5, at least 6, and at least 7. , At least 8 types, at least 9 types, at least 10 types, at least 11 types, at least 12 types, at least 13 types, at least 14 types, at least 15 types, at least 20 types, at least 25 types, at least 30 types, at least 35 types, at least The differential expression of 40, at least 45, or 46 miRNAs can be measured and at least 2 of the miRNAs listed as "down-regulated" in Table 11 or Table 17. Types, at least 3 types, at least 4 types, at least 5 types, at least 6 types, at least 7 types, at least 8 types, at least 9 types, at least 10 types, at least 11 types, at least 12 types, at least 13 types, at least 14 types, At least 15, at least 20, at least 25, or 28 differential expressions can be measured.

In one example, the methods described herein that provide a diagnosis or indicator of whether a subject may have stage 4 gastric cancer use Table 17. The method expresses, but is not limited to, expression of any one or more miRNAs, including, but not limited to, miR-21-5pm, miR-223-3pm, miR-20a-5pm, miR-106b-5p, and miR-17-5p. An additional step of comparing levels may be included.

Yet another aspect provided is a method of detecting or diagnosing gastric cancer in a subject, or a method of determining the likelihood that a subject has gastric cancer, in (a) a non-cellular biofluid sample. The steps of measuring the expression level of at least one miRNA having at least 90% sequence identity with the miRNA listed in Table 12 or Table 19 and (b) the expression level of the miRNA measured in step (a). A step of generating a score based on the subject and (c) a step of predicting the likelihood that the subject has gastric cancer using the score are included, and the expression level of the subject is positively controlled (derived from a gastric cancer subject). The score is calculated by a classification algorithm that compares it with a sample) or a negative control (a sample from a subject without gastric cancer), and the score is such that the subject is within the range of (i) the score of the positive control. A method of identifying the likelihood of either having gastric cancer that fits; or (ii) the score falls within the negative control score, or has no gastric cancer (ie, no gastric cancer). In one example, the miRNA can include at least one of the miRNAs listed in Table 19. In one example, a method of detecting or diagnosing a subject's gastric cancer, or a method of determining the likelihood that a subject has gastric cancer, may be referred to as a multivariate method. In one example, the multivariate method may use miRNAs that are representative of subpopulations or groups as determined by clinical data. In one example, the multivariate biomarker panel contains the biomarkers listed in Table 12 (containing both miRNAs known to be associated with gastric cancer and miRNAs known not to be associated with gastric cancer). May be included. In one example, miRNAs that can be used in multivariate methods and are associated with gastric cancer may be included in Table 19. Table 12 is described in more detail in the Experimental section. In one example, FIG. 14 summarizes the steps of a multivariate method that determines the likelihood that a subject will develop (affect) gastric cancer.

Table 19 lists miRNAs that are often selected for use in multivariate biomarker panels with altered miRNA expression levels in gastric cancer subjects, which are:

As mentioned above, the multivariate method is the step of determining the biomarkers listed in Table 12 (including both miRNAs known to be associated with gastric cancer and miRNAs known not to be associated with gastric cancer). May include. Thus, in one example, using miRNAs whose multivariate method is listed in Table 19, the method measures the expression level of at least one miRNA whose expression level has not changed in the subject when compared to the negative control. May include additional steps. In one example, the miRNA whose expression level has not changed in the subject when compared to the negative control may be any one miRNA listed in Table 20.

Table 20 lists miRNAs that are often selected for use in multivariate biomarker panels for which miRNA expression levels were not associated with gastric cancer subjects, and are as follows:

In one example, in the multivariate methods known herein, the scores can be calculated using a classification algorithm. In one example, the classification algorithms are, but are not limited to, support vector machine algorithms, logistic regression algorithms, polynomial logistic regression algorithms, Fisher's linear discriminant algorithms, quadratic classification algorithms, perceptron algorithms, k-neighborhood methods, artificial neural circuits. Includes network algorithms, random forest algorithms, decision tree algorithms, naive Bayes algorithms, adaptive Bayes network algorithms, and ensemble learning methods that combine multiple learning algorithms. As will be appreciated by those skilled in the art, classification algorithms can be pretrained using expression levels of positive and negative controls.

From a particular point of view of the multivariate method described herein, the positive control may be a sample obtained from a subject with gastric cancer and the negative control may be a sample obtained from a control without gastric cancer. May be good. In one example, gastric cancer-free controls may be those described above. That is, in one example, a gastric cancer-free control may be a subject who has no gastric cancer-related disease. In one example, gastric cancer-free controls may be healthy individuals with no history. In one example, a gastric cancer-free control may be a subject known to have a non-gastric cancer-related disorder (eg, but not limited to, gastritis, intestinal metaplasia, intestinal atrophy, etc.). In one example, the controls may vary in sample properties and other factors (eg, subject gender, age, race, based on the subject from whom the training dataset was obtained). In one example, the control subject may be a Chinese or a descendant of a Chinese.

As will be appreciated by those skilled in the art, the expression level of miRNA may be any one of concentration, log (concentration), Ct / Cq value, 2 Ct / Cq power value, and the like.

In one example, the methods described herein may use data-trained formulas that may be able to determine the likelihood that a subject has gastric cancer. In one example, the algorithm combines multiple miRNA measurements to give a predictive score for a subject's risk of having gastric cancer. The algorithm combines information from multiple gene targets to give better predictive accuracy than using a single miRNA biomarker. In one example, the minimum number of miRNA biomarkers for constructing useful panels is four. For example, as shown in FIG. 11, a biomarker panel with four miRNAs has an average AUC of> 0.8 in a validation set that would be a useful biomarker panel. The improvement in biomarker panel performance (in terms of AUC value) does not increase significantly when the number of miRNA types is greater than 7. Therefore, in one example, the minimum number of miRNAs to build the optimal panel with maximum performance is 7 types of miRNAs. In one example, the formulas that can be used to calculate the score for determining the likelihood that a subject has gastric cancer are:

A x miRNA1 + B x miRNA2 + C x miRNA3 + D.

In the formula, miRNA1, miRNA2, and miRNA3 are values associated with the expression level of miRNA, which may be copy number, log (copy number), Ct / Cq value, or other value. Then, A, B, C, and D are coefficients, which may be positive or negative. For example, a score (ie, a cancer risk score) for determining the likelihood that a subject has gastric cancer is often selected in the multivariate biomarker panel identification process, as described in the examples below and FIG. It is possible to calculate by combining 12 types of miRNAs (Table 21) having an appearance rate of> 20%. In one example, when the score is obtained from the equation, the higher / lower the score, the higher the probability that the subject will have gastric cancer. In one example, the range may not be clearly defined because the coefficients may be adjusted. For example, all coefficients (ie, A, B, or C) may be halved, or the value of D may vary depending on training with the training dataset. In one example, the score value may be positively or negatively correlated with the risk of having gastric cancer. The actual correlation may depend on the training dataset, and those skilled in the art are familiar with the necessary adjustments required in establishing the training dataset.

In one example, the range may not be clearly defined because the coefficients may be adjusted. For example, all coefficients (ie, A, B, or C) may be doubled, or the value of D may vary depending on training with the training dataset. In one example, the score value may be positively or negatively correlated with the risk of having gastric cancer. The actual correlation may depend on the training dataset, and those skilled in the art are familiar with the necessary adjustments required in establishing the training dataset.

In one example, all coefficients (ie, A, B, or C) can be deliberately prepared in half, and / or the value of D can also be deliberately modified for most subjects. The range of values calculated based on the equation can be adjusted from 0 to 100. In one example, the score value may be positively or negatively correlated with the risk of having gastric cancer. The actual correlation may depend on the training dataset, and those skilled in the art are familiar with the necessary adjustments required in establishing the training dataset.

In one example, the methods described herein are at least one miRNA listed as "significant" in Table 12 or Table 19 and at least listed as "non-significant" in Table 12 or Table 20. The expression level of one miRNA can be measured. In one example, the method is listed as "significant" in Table 12, at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, and at least 9. Differences between at least 10 types, at least 11 types, at least 12 types, at least 13 types, at least 14 types, at least 15 types, at least 20 types, at least 25 types, at least 30 types, at least 35 types, or 42 types of miRNA. Expression can be measured, or at least two, at least three, at least four, at least five, at least six, at least seven, at least eight, at least the miRNAs listed in Table 19. 9 types, at least 10 types, at least 11 types, at least 12 types, at least 13 types, at least 14 types, at least 15 types, at least 20 types, at least 25 types, at least 30 types, at least 35 types, or 38 types Is expression measurable; and at least two, at least three, at least four, at least five, at least six, at least seven, at least miRNAs listed as "non-significant" in Table 12. Eight, at least nine, or twelve differential expressions can be measured, or at least two, at least three, at least four, at least five miRNAs listed in Table 20. At least 6, at least 7, at least 8, at least 9 or 10 differential expressions can be measured.

In one example, if the multivariate methods described herein use gastric cancer-related miRNAs listed in Table 19, the methods are miR-21-5p, miR-20a-5pm, miR-17-. It may further include measuring the expression level of any one or more miRNAs selected from the group consisting of 5p, miR-423-5p, and miR-223-3p.

In one example where the multivariate methods described herein use unrelated miRNAs listed in Table 20, the methods described herein are from miR-34a-5p and miR-27b-3p. Further may include the step of measuring the expression level of any one or more miRNAs selected from the group.

An exemplary approach to conceive of miRNA sets that correlate with gastric cancer is summarized below. The general work flow is also shown in FIGS. 1, 2, and 14.

Step 1: Total RNA (or its subfractions) from a sample (eg, plasma, serum, or other blood fraction) of a subject (s) with gastric cancer using the appropriate kit and / or purification method. And extract.

Step 2: From each sample, the amount (expression level) of one miRNA or at least one set of miRNAs selected from the group consisting of SEQ ID NO: 1 to SEQ ID NO: 191 using experimental techniques. Measure. In one example, at least one miRNA can be selected from the groups listed in Tables 12, 19, and 20. These techniques are known in the art and are not limited to array-based approaches, amplification methods (PCR, RT-PCR, or qPCR), sequencing, next-generation sequencing, and / or mass. Includes analysis.

Step 3: Mathematical methods are applied to collect diagnostic values and duplications for each single miRNA biomarker. These methods are known in the art and are not limited to basic mathematical approaches (eg, found quotients, signal-noise ratios, correlations), statistical methods (hypothesis testing (hypothesis testing (hypothesis testing)). (Example, t-test, Wilcoxon-Mann-Whitey test)), area under the receiver motion characteristic curve, information theory approach (eg, mutual information amount, cross-entropy), probability theory (eg, coupling probability and conditional probability) ), Or a combination and modification of the methods described so far.

Step 4: Use the information obtained in Step 3) to infer diagnostic content or diagnostic values for each miRNA biomarker. However, this diagnostic value is usually too small to obtain a very accurate diagnosis with an accurate rate, specificity, and sensitivity above the 80% barrier. The contents of the diagnosis of miRNA suitable for the diagnosis of gastric cancer are shown in FIG. This figure includes miRNAs listed in Tables 12, 19, and 20.

Step 5: More than one miRNA biomarker may be employed to improve the diagnostic outcome of individuals with gastric cancer. Therefore, statistical learning / machine learning to select / define tailored miRNA biomarker sets (which may contain miRNAs listed in Tables 12 or 19 and 20) for the detection of gastric cancer. / Bioinformatics / Apply a computer approach to set selection. These methods are not limited, but are described in the wrapper subset selection method (eg, forward stepwise method, backward stepwise method, combination method, optimization method) and filter subset selection method (eg, step 3). Methods), principal component analysis, or combinations and modifications of such methods (eg, hybrid methods).

Step 6: The subset selected / defined in Step 5 (which may be in the range of only a few (at least one, at least two, or more sets) to all measured miRNAs) is subsequently used. And make a diagnosis of gastric cancer. To this end, we apply statistical learning / machine learning / bioinformatics / computerized approaches to, but not limited to, analysis with or without objectives of any type (classification). Method (eg, naive bayes method, linear discriminant analysis, neural network for secondary discriminant analysis, tree-based approach, support vector machine, neighborhood method), regression method (eg, linear regression, multiple regression, logistic regression, probit regression, Sequential logistic regression, ordinal probit regression, Poisson regression, negative binary regression, polynomial logistic regression, cutting regression), clustering method (eg, k-averaging method, hierarchical clustering, PCA), modification and extension of the approaches described so far , And combinations) are included.

Step 7: By combining the subset selection (step 5) and the machine learning approach (step 6), an algorithm or mathematical function for diagnosing gastric cancer is obtained. This algorithm or mathematical function is applied to an individual (subject) miRNA profile (miRNA expression profile data) to make a diagnosis of gastric cancer.

In any of the methods described herein, the non-cellular biological fluid may include any body fluid that does not contain any cellular components of the body fluid. Thus, in one example, when extracted from a subject, the body fluid or body fluid may contain cellular components commonly found in the subject's body fluid or body fluid. However, the biofluid or body fluid can also be treated using methods known in the art to remove any cellular component contained in the biofluid or body fluid. For example, a centrifugation or purification process may be performed to remove cellular components of body or biological fluids, after which the methods of the present disclosure may be performed. Thus, in one example, body fluids or non-cellular biofluids are, but are not limited to, sheep water, breast milk, bronchial lavage fluid, cerebrospinal fluid, primary milk, interstitial fluid, peritoneal fluid, pleural fluid, saliva, semen, urine. , Tears, non-cellular components of whole blood (including plasma, red blood cells, white blood cells, and serum). In one example, the non-cellular biofluid may be serum or plasma.

In any of the methods described herein, the terms "subject" and "patient" as used herein are used interchangeably with animals (eg, mammals, fish, amphibians, reptiles, etc.). Can refer to birds, and insects). In one example, the subject may be mammals (eg, non-human mammals (eg, dogs, cats, or primates) and humans). In one example, the subject may be a primate (eg, chimpanzee and human). In one example, the subject may be human. In one example, the subject may be a human with or without gastric cancer (stomach cancer). In the experimental section of the disclosure, training datasets were obtained from subjects who were Chinese or descendants of Chinese. Therefore, in one example, the subject may be an Asian descendant. In one example, the subject may be a Chinese or a descendant of a Chinese.

In one example, the methods described herein can be provided as a kit. The kit may include the required components or components that detect the miRNAs described herein. Components include primers / probes, reagents, containers, and / or devices configured to detect the miRNAs and primers / probes described herein that are hybridizable with the miRNAs described herein. May be included. In one example, the container in the kit contains a primer / probe set capable of detecting or measuring the levels of miRNA described herein, which is labeled with a radioactive element prior to use. There is. The kit contains a container of material for reverse transcription and amplification of one or more miRNAs (eg, dNTPs (4 different deoxynucleic acids dATP, dCTP, dGTP, and dTTP), buffers, reverse transcriptase, and DNA polymerase. Multiple containers including) may be included. The kit may also include nucleic acid templates (s) for positive and / or negative control reactions. The kit may further include any reaction component or buffer required to isolate the sample-derived miRNA obtained from the subject.

The invention described explanatoryly in the present specification can be appropriately carried out without any element (s) and restrictions (s) not specifically disclosed in the present specification. Thus, for example, the terms "comprising," "inclusion," "contining," etc. should be read broadly and without limitation. In addition, the terms and expressions adopted in the present specification are used as explanatory terms and are not intended to be limited. And there is no intention to use terms and expressions that exclude any equivalent of the features shown and described or parts thereof. However, it is recognized that various modifications are possible within the scope of the claimed invention. Therefore, it should be understood that while the invention has been specifically disclosed by preferred embodiments and optional features, modifications and variations of the invention materialized herein and elsewhere are those skilled in the art. It is recoverable by, and such modifications and modifications are considered to be within the scope of the present invention.

The present invention has been broadly and generally described herein. Narrower species and subgenus groups within the general disclosure also form part of the invention. This is a general description of the invention with hypothetical or negative limitations that remove any subject element from its genus, whether or not the material taken up is specifically described herein. Including.

Other embodiments are within the following claims and non-limiting examples. Also, if a feature or aspect of the invention is described in Markush group terms, those skilled in the art will appreciate that the invention is also described thereby with respect to any individual member of the Markush group or a subgroup of members. Recognize.

Experimental section Method Analysis pretreatment (sample collection and microRNA extraction): Serum samples from normal subjects (ie, gastric cancer-free controls) and gastric cancer subjects were centrifuged, collected, processed, and processed by members of the Singapore Gastric Cancer Association. Then, it was cryopreserved at −80 ° C. 200 μl of total serum-derived RNA was isolated on a semi-automated QiaCube system using a well-established column-based miRNeasy kit (Qiagen, Hilden, Germany). Serum contains a small amount of RNA. Reasonably designed isolation enhancer (MS2) and spike-in control RNA (MiRXES) were added to the specimen prior to isolation to reduce RNA loss and monitor extraction efficiency.

RT-qPCR: With total RNA isolated in an optimized multiple reverse transcriptase with a second set of spike-in control RNA to detect the presence of inhibitors and monitor RT-qPCR efficiency. Synthetic RNA standards were converted to cDNA. The reverse transcription reaction was performed using Impron II (Promega®) reverse transcriptase according to the manufacturer's instructions. The synthetic cDNA was then subjected to a multiplex enhancement step and quantified using the SYbr Green-based single qPCR assay (MIQE compatible) (MiRXES®). The ABI® ViA-7 384 device was used for the qPCR reaction. The outline and details of the work flow for RT-qPCR measurement of miRNA are summarized in FIG.

Data processing: The number of cycles to the unprocessed threshold (Ct) was processed and the absolute copy number of the target miRNA in each sample was determined by interpolating the calibration curve of the synthetic miRNA. The technical variability introduced during RNA isolation and RT-qPCR processing was standardized by spike-in control RNA. For single miRNA analysis, biological variability was further standardized by a validated endogenous reference miRNA set that was stably expressed across all controls and disease samples.

Study Design A well-designed clinical trial (case control study) was conducted to ensure accurate identification of diagnostic biomarkers for gastric cancer. A total of 237 Chinese patients with gastric cancer (stages 1-4) with an average age of 68.0 ± 10.9 years were used in this study. They were then compared to another 236 normal (healthy / gastric cancer-free) Chinese subjects of age and gender who functioned as a control group. All cancer subjects were confirmed by endoscopy and biopsy, and serum samples were collected prior to any treatment. All normal (healthy) subjects were confirmed by endoscopy at sample collection for gastric cancer, and were screened for endoscopy every two years and followed up for 3-5 years. It was confirmed that gastric cancer did not develop within the period. Detailed clinical information of the subjects is tabulated in Tables 2 and 3. All serum samples were stored at −80 ° C. before use. The characteristics of the subjects tested are summarized in Table 5. A large number of control subjects (normal / healthy / non-gastric cancer subjects) had gastric disorders (eg, gastritis, intestinal metaplasia, and intestinal atrophy). Therefore, the control group corresponded to the high-risk population for gastric cancer. The percentage of patients with gastric cancer who have H. pylori (78.8%) is higher than that of the control group (55.7%), but the difference is small, so we support the use of H. pylori as a biomarker for gastric cancer. Was not enough.

It is known that cell-free miRNAs that circulate in the blood may originate from a variety of tissue sources. As a result, changes in miRNA levels due to the presence of solid gastric tumors may be complicated by the presence of the same miRNA from other sources. Therefore, it is difficult to determine the difference in miRNA expression levels found in cancer and the control group, and the difference is not as clear as expected. In addition, it is known that most of the miRNA released from cells is extremely small in blood due to the diluting effect due to the large volume of blood (5 L in adults). Therefore, accurate measurement of multiple miRNA targets from a limited amount of serum / plasma samples is important and presents a very large challenge. Instead of using low-sensitivity or semi-quantitative screening methods (microarrays, sequencing), it is easiest to discover significant changes in miRNA expression and identify multivariate miRNA biomarker panels for the diagnosis of gastric cancer. Therefore, in this experimental study, we chose to perform a qPCR-based assay using the workflow summarized in FIG.

In the workflow used in this experimental section, all reactions were performed at least twice in a single mode for miRNA targets and at least four times for synthetic RNA "spike-in" controls. To ensure the accuracy of the results of such high-throughput qPCR tests, after repeated times, the test was designed / established into a powerful workflow for discovering circulating biomarkers (“Methods”). See section and Figure 2). In this workflow, various artificially designed "spike-in" controls were used to monitor and correct technical variability in isolation, reverse transcription, enhancement, and qPCR processes. All spike-in controls are unnaturally synthesized miRNA mimetics (small molecule single-stranded RNAs ranging in length from 22 to 24 bases), and they are incilico such that they have very low sequence homology to all known human miRNAs. Designed in, therefore, cross-hybridization to the primers used in the assay was minimized. In addition, the miRNA assay was devised and divided into a large number of multiplex groups by in silico to minimize non-specific amplification and primer-primer interaction. Synthetic miRNAs were used to create calibration curves for absolute copy number interpolation in all measurements and thus further corrected for technical variability. When using this workflow with multiple levels of control, it is expected that even circulating miRNAs with low expression levels can be reliably and reproducibly detected.

miRNA biomarkers In the process of identifying biomarkers, the expression levels of each miRNA in the normal (control) and diseased states are compared. Expression levels of human miRNAs in all 578 serum samples (gastric cancer and normal / control) (miRBase version 10 release) using a strong workflow and a sensitive qPCR assay (MiRXES, Singapore) It was measured quantitatively.

In this experimental design, 200 μl of serum was extracted, total RNA was reverse transcribed, and enhanced by touchdown amplification to increase the amount of cDNA, but not the ratio of miRNA expression levels (Fig. 2). The enhanced cDNA was then diluted for qPCR measurement. A simple calculation based on the dilution effect revealed that miRNAs (expressed in serum at the level of ≤500 copies / ml) are close to the detection limit of the single qPCR assay (≤10 copies / well). It is to be quantified. At such concentrations, the measurement becomes very difficult due to technical limitations (errors in pipetting and qPCR). Therefore, miRNAs expressed at a concentration of ≤500 copies / ml were excluded from the analysis and considered undetectable in this study.

Approximately 33% of the total assayed miRNAs (total of 578 miRNAs assayed) were found to be highly expressed. These 191 miRNAs (expression level ≥ 500 copies / ml; Table 4) were reliably detected in more than 90% of serum samples. And these were much more than previously reported using other technologies. This underscores the importance of using a strong experimental design and a well-controlled workflow.

Heatmaps were then created to represent the expression levels of all detected serum miRNAs (191 detected miRNAs, FIG. 3). Although miRNAs from various subpopulations were clustered together, there did not appear to be a clear difference between the cancer and control groups due to hierarchical clustering without objective variables based on the total number of miRNAs detected. Therefore, without any feature selection, the overall miRNA profiles for both groups looked similar.

The expression of 191 serum miRNAs was then compared.

A] Between normal / control and gastric cancer (regardless of stage or subtype).
-B] Between various stages of gastric cancer (stages 1 to 4), and-C] Between normal / control and various stages of gastric cancer.

The significance of the differential expression between the two groups is a t-test (p-value <0.01) corrected for false discovery rate (FDR) estimates using the Bonferroni-type multiple comparison method known in the art. Calculated based on.

A] Identification of miRNAs differentially expressed in normal / control and gastric cancer (regardless of stage or subtype) Clinically confirmed to have either diffuse, intestinal, or mixed gastric cancer Patient-derived samples were grouped together and compared to normal / control donor-derived samples. A pool of 75 miRNAs showing significant differential expression between the control and cancer groups was identified (p-value <0.01, Table 6). Unlike other reports (Table 1), where most of this putative biomarker expression was found to be higher in gastric cancer subjects, this study up-regulated 51 miRNAs in cancer subjects. Although rated, 24 were demonstrated to be down-regulated (Table 6). Therefore, this experimental design identified more regulated biomarkers.

Comparing the results of this study (Table 6) with the 30 miRNAs (Table 1) identified in other reports as differentially expressed in blood, only 6 miRNAs (hsa-miR-106b) -5p, hsa-miR-17-5p, hsa-miR-20a-5p, hsa-miR-21-5p, hsa-miR-223-3p, and hsa-miR-423-5p) are all up. It was found to be regulated (Table 7). Therefore, most of the miRNAs that are said to be differentially regulated in the literature were not confirmed in this study. Interestingly, 69 novel miRNAs, which are potential biomarkers for gastric cancer, have been identified here.

Using this set of 75 biomarkers, more differential clustering between gastric cancer subjects and normal / control subjects was observed in the heatmap of the miRNA profile (FIG. 4). Looking at the horizontal lines, the majority of gastric cancer subjects (black) were clustered in one concentrated area, and the majority of normal / control subjects were left in the remaining space in the map. The AUC values for the top-ranked up-regulated miRNA (hsa-miR-142-5p) and down-regulated miRNA (hsa-miR-99b-5p) were only 0.71 and 0.67, respectively. There was (Fig. 5). Each or a combination of several of the 75 differential miRNAs can serve as a biomarker or biomarker panel for the diagnosis of gastric cancer.

The miRNA expression was found to be clustered into subgroups as described in the heatmaps (FIGS. 3 and 4). Each cluster has about 10 to 20 types of miRNA (FIGS. 3, 4; between dotted lines). Analysis of all detectable miRNAs revealed that many of these were positively correlated (Pearson correlation coefficient> 0.5) (Fig. 6). The same observations were found for gastric cancer-specific miRNAs (Fig. 7). These results showed that a group of miRNAs were similarly regulated among all subjects. As a result, a panel of miRNAs can be assembled by substituting one or more different miRNAs with another to improve diagnostic outcomes. And all 75 significantly altered miRNAs were important in the development of multivariate index diagnostic assays for gastric cancer.

B] Identification of miRNAs that are differentially expressed between normal / control and various stages of gastric cancer.
The expression of miRNAs between the various stages of gastric cancer was then compared. With bidirectional ANOVA taking into account stages and subtypes, 36 miRNAs (derived from a pool of 191 serum miRNAs) can be expressed differentially between different stages of gastric cancer subjects. It was found (the p-value of the ANOVA test after false discovery rate correction was <0.01. Data not published).

Knowing the differences between the different stages, the next step was to identify the miRNAs that are differentially expressed between normal / control and each stage of cancer. The total number of different miRNAs found during this new search process was 113 (Tables 8-11; in those tables, the profiles were duplicated). Among these, 20 types of miRNA (8 types of up-regulated and 12 types of down-regulated), 62 types of miRNA (54 types of up-regulated and 8 types of down) 43 types of miRNA (37 types of up-regulated and 6 types of down-regulated), or 74 types of miRNA (28 types of up-regulated) Pools of 46 down-regulated (46 types of down-regulated) showed differential expression between normal / control and the corresponding stages 1, 2, 3, and 4, respectively (after FDR correction). The p-value is <0.01; Tables 8-11).

With the exception of three types of miRNAs (hsa-miR-486-5p, hsa-miR-19b-3p, hsa-miR-411-5p), they were discovered from analyzes that use all stages of cancer together. The rest of the 75 miRNAs (Table 6) were found to be subsets of these 113 miRNAs identified during independent analysis of each stage of cancer. Most of the 36 miRNAs found to be differentially regulated between different stages of gastric cancer (except one) compared normal / control (N) with each stage of gastric cancer (S1-S4). It was found that it was expressed differentially (Tables 8 to 11). Differences between each stage resulted in different miRNA pools that were useful in detecting gastric cancer. Therefore, the results of comparing expression levels at different stages to normal / control resulted in a larger pool of beneficial miRNAs.

Apparently, there were fewer miRNAs identified in early stage 1 than in late stages. The features of miRNAs with the highest AUC values were not always the same at all stages (Fig. 8). A larger number of miRNAs were regulated (Tables 8-11). Then, similar features were repeatedly found over a plurality of stages (Fig. 10). It was found that only one miRNA (hsa-miR-27a-5p) was common among all four stages. These different miRNAs (Tables 8-11) can be used individually, and collecting the results of multiple miRNAs can be used to diagnose gastric cancer in clinical settings where the stage of cancer is unknown at the time of examination. May be useful.

From the above studies, a biomarker panel consisting of at least one miRNA from a list of miRNAs specific for each stage can provide an alternative method of determining the likelihood that a subject will have gastric cancer.

III Search for multivariate biomarker panels An important endpoint for assembling multivariate panels covers all cancer subgroups, including at least one miRNA from a specific list for each stage of cancer. It is to guarantee that. However, the miRNAs that define the four stages of cancer were not completely different. This is because the same miRNAs are found between those stages as well (Fig. 9). At the same time, many cancer-related miRNAs and unrelated miRNAs were found to be positively correlated (FIGS. 6 and 7), which makes it difficult to select the optimal miRNA combination for gastric cancer diagnosis.

Given the complexity of this task, we decided that in this study it was necessary to identify the miRNA panel with the highest AUC using the SFFS (sequence forward floating sequence) algorithm. It is also known to use state-of-the-art linear support vector machines (a well-used and recognized modeling tool for building variable panels) in the field to assist in the selection of miRNA combinations. It was. This model produces a score based on a linear equation that describes the expression level of each member and its weighted value. This linear model is easily accepted and applied in clinical practice.

A key requirement for the success of such a process is the availability of quality data. Quantitative data of all detected miRNAs in a large number of well-defined clinical samples have not only improved the accuracy and accuracy of the results, but have also been identified for clinical applications using qPCR. It also ensured the consistency of the biomarker panel.

In a large number of clinical samples (500+), the problem of over-fitting data in modeling was minimized. That's because there are only 191 feature candidates to choose from. Also, to ensure the accuracy of the results, quadruple cross-validation was performed multiple times and an independent set of validation samples (remaining samples at each weight) based on the discovery set (3/4 of the sample at each weight). The performance of the identified biomarker panel in 1/4) was validated. During the cross-validation process, samples were fitted with respect to gender, subtype, and stage.

A boxplot of the results (AUC of both discovery and verification biomarker panels) is shown in FIG. As expected, there was a decrease in the AUC value of the validation set for each search (0.04-0.06). The size of the box was larger in the validation phase data (indicating a wider range of values), but the difference in AUC values was ≤0.1. The breadth of this value range was even narrower when the number of biomarkers included in the analysis was greater than 6 (<0.05). This observation showed that stronger results were obtained with miRNA numbers of 6 and above.

A more quantitative representation of the results is shown in FIG. There, no improvement in AUC value was observed when the number of biomarkers included in the analysis was greater than 7. The difference between the biomarker panel consisting of 6 miRNAs and the biomarker panel consisting of 7 miRNAs was statistically significant, but the improvement in AUC value was only 0.02. Therefore, an AUC value of 0.855 obtained using a panel of 6 miRNAs is sufficient for clinical use.

Composition of IV multivariate miRNA biomarker panels To investigate the composition of multivariate biomarker panels, the abundance of miRNAs in all panels containing 6-10 miRNAs was counted, with the top 10% and bottom 10%. Panels with AUC were excluded. Count false-discovery biomarkers resulting from applying inaccurate data from subpopulations generated by a randomized process of cross-validation analysis, excluding those with top 10% and bottom 10% AUC. I avoided that. When excluding these miRNAs when the count was less than 4, a total of 55 miRNAs were selected during the discovery process (Table 12). It was also found that the expression of 43 of these was significantly changed in cancer. Including the other 12 types (although there was no change in cancer) was found to significantly improve the AUC value. Without direct and quantitative measurement of all miRNA targets, these miRNAs would never have been selected by high-throughput screening methods (microarrays, sequencing) and were excluded for qPCR validation. There will be. These miRNAs are not randomly selected. This is because the top two miRNAs (hsa-miR-532-5p, hsa-miR-30e-5p) were consistently found in 57.8% and 48.4% of these panels.

MiRNAs selected to form multivariate panels (Table 12) showed differences in detecting different stages of cancer (Tables 8-11). For the top 10 frequently selected miRNAs listed as having a prevalence of more than 20%, only 3 of them (hsa-miR-21-5p, hsa-miR-27a-) Only 5p, hsa-miR-142-5p) were found to be commonly controlled in more than two different stages of cancer. On the other hand, for the rest, hsa-miR-103a-3p is significant only in stage 1, hsa-miR-181a-5p is significant only in stage 4, and hsa-miR-26a-5p is significant in stages 1 & 4. Hsa-miR-20a-5 and hsa-miR-17-5p are significant for stages 2 & 4, and hsa-miR-616-5p and hsa-miR-30a-5p are significant for stages 3 & 4. Met. Two types of miRNAs specific for each of stages 1-3 were often selected, excluding stage 4 (FIG. 9), where a large number of miRNAs overlapped the list of the other three stages.

When comparing the identities of multiple miRNAs selected for multivariate panels as diagnostic markers with a single miRNA, they were not necessarily the same. For example, both the top up-regulated miRNA (hsa-miR-532-5p) and the down-regulated miRNA (hsa-miR-30e-5p) for all cancers are listed. Was not included in. Therefore, it was not possible to simply combine the best biomarkers identified for different stages to form the optimal biomarker panel, but rather the marker panel provided complementary information that gave the best results.

Interestingly, five miRNAs (those with> 48% appearance rate, see Table 12) were often selected for multivariate panels. Three of these (hsa-miR-21-5p, hsa-miR-103a-3p, and hsa-miR-20a-5p) are cancer-specific, and two (non-significant group; hsa-). miR-532-5p and hsa-miR-30e-5p) were derived from other groups. The rate at which these five miRNAs were present simultaneously in all multivariate panels was analyzed (Table 13). Then, 85.9% of the panels had either hsa-miR-532-5p or hsa-miR-30e-5p. Therefore, the use of any one of these two miRNAs may be sufficiently useful to increase the AUC value in a multivariate index panel.

Overall, 93.8% of the miRNA biomarker panels had at least one miRNA from each group. Thus, in one example, the multivariate method of determining the likelihood that a subject has gastric cancer may include at least one miRNA from each of these two often selected miRNA groups.

Calculation of Cancer Risk Scores MiRNAs can be combined to form biomarker panels to calculate cancer risk scores (Equation 1). For example, a biomarker panel is formed by combining 12 miRNAs (Table 21) that have an incidence of> 20% and are often selected in the multivariate biomarker panel identification process to calculate the cancer risk score. can do. The formula here demonstrates the use of a linear model for gastric cancer risk prediction, and the cancer risk score (unique for each subject) indicates the likelihood that a subject will have gastric cancer. This calculation is performed by adding the weighted values of 12 types of miRNA and the constant 50.
Equation 1-

log ₂ copy number_miRNA _i- 12 log-converted copy numbers of individual miRNAs. _Ki − Coefficients used to weight multiple miRNA targets, as seen in Table 21. The value of K _i, and optimized using support vector machines method was adjusted to the range of 0-100. Subjects with a cancer risk score lower than 0 are considered to be 0, and subjects with a cancer risk score higher than 100 are considered to be 100.

The control subject and the cancer subject in this study have different cancer risk score values calculated based on the above formula (Fig. 15A). A match of the cancer risk scores of the control subject and the cancer subject to the probability distribution can be seen in FIG. 15B, and it can be found that there is a clear separation between the two groups. In this study, the control subjects were non-cancer subjects selected from the high-risk population. And the probability of having gastric cancer is 0.0067. Based on this prior probability and the one adapted to the probability distribution of FIG. 15B, the probability (risk) of an unknown subject having cancer can be calculated based on their cancer risk score values (FIG. 15C). ). The higher the score, the higher the subject's risk of having stomach cancer. Furthermore, when compared with the incidence of gastric cancer in the high-risk population, the cancer risk score can be used to determine the rate of change in the probability (risk) of an unknown subject having gastric cancer (FIG. 15D). For example, an unknown high-risk subject with a cancer risk score of 70 has a 14.6% chance of having gastric cancer, which is about 22 times higher than the average risk of the high-risk population.

By further applying the two thresholds 40 and 70, subjects can be defined into three categories based on cancer risk scores (FIGS. 15C, D). Subjects with a cancer risk score in the range 0-40 belong to a group with a low cancer risk, and that group has a 0.052% chance of having gastric cancer. This is 13 times lower than the average cancer incidence in the high-risk population. Subjects with a cancer risk score in the range of 40-70 belong to a group with a moderate cancer risk, and that group has a 1.4% chance of having gastric cancer. This is twice as high as the average cancer incidence in the high-risk population. Subjects with a cancer risk score in the range of 70-100 belong to a group with a high risk of cancer, and that group has a 12.2% chance of having gastric cancer. This is 18 times higher than the average cancer incidence in the high-risk population.

Claims (23)

A method of detecting or diagnosing a subject's gastric cancer, or a method of determining the likelihood that a subject will develop gastric cancer.
It comprises the step of measuring the expression level of at least one miRNA having at least 90% sequence identity with the miRNA in the non-cellular biofluid sample obtained from the subject.
The differential expression of the miRNA expression when compared to the control serves as an indicator for the subject with gastric cancer.
The miRNAs are hsa-miR-142-5p, hsa-miR-29c-3p, hsa-miR-93-5p, hsa-miR-140-5p, hsa-miR-148a-3p, hsa-miR-183-. 5p, hsa-miR-29b-3p, hsa-miR-424-5p, hsa-miR-101-3p, hsa-miR-106b-3p, hsa-miR-128, hsa-miR-1280, hsa-miR- 140-3p, hsa-miR-15b-3p, hsa-miR-186-5p, hsa-miR-18b-5p, hsa-miR-197-3p, hsa-miR-19a-3p, hsa-miR-19b- 3p, hsa-miR-20b-5p, hsa-miR-21-3p, hsa-miR-23a-5p, hsa-miR-25-3p, hsa-miR-27a-5p, hsa-miR-29a-3p, hsa-miR-29b-2-5p, hsa-miR-29c-5p, hsa-miR-338-5p, hsa-miR-425-3p, hsa-miR-4306, hsa-miR-450a-5p, hsa- miR-486-5p, hsa-miR-500a-3p, hsa-miR-501-5p, hsa-miR-532-3p, hsa-miR-550a-5p, hsa-miR-579, hsa-miR-589- 5p, hsa-miR-590-5p, hsa-miR-598, hsa-miR-616-5p, hsa-miR-627, hsa-miR-629-3p, hsa-miR-629-5p, hsa-miR- 93-3p, hsa-miR-195-5p, hsa-miR-18a-3p, hsa-miR-363-3p, hsa-miR-181a-2-3p, hsa-miR-16-5p, hsa-miR- 501-3p, hsa-miR-23a-3p, hsa-miR-339-3p, hsa-miR-15a-5p, hsa-miR-320b, hsa-miR-374b-5p, hsa-miR-650, hsa- miR-1290, hsa-miR-22-3p, hsa-miR-320c, hsa-miR-130a-3p, hsa-miR-320e, hsa-miR-378a-3p, hsa-miR-9-5p, hsa- miR-200b-3p, hsa-miR-141-3p, hsa-miR-191-5p, hsa-miR Is it an "up-regulated" miRNA selected from the group consisting of -628-5p, hsa-miR-484, hsa-miR-425-5p; or
hsa-miR-103a-3p, hsa-miR-30a-5p, hsa-miR-181a-5p, hsa-miR-107, hsa-miR-26a-5p, hsa-miR-126-3p, hsa-miR- 99b-5p, hsa-miR-339-5p, hsa-miR-122-5p, hsa-miR-136-5p, hsa-miR-139-5p, hsa-miR-146a-5p, hsa-miR-154- 5p, hsa-miR-193b-3p, hsa-miR-23c, hsa-miR-30b-5p, hsa-miR-337-5p, hsa-miR-382-5p, hsa-miR-409-3p, hsa- miR-411-5p, hsa-miR-485-3p, hsa-miR-487b, hsa-miR-495, hsa-miR-885-5p, hsa-miR-99a-5p, hsa-miR-362-5p, hsa-miR-671-3p, hsa-miR-454-3p, hsa-miR-328, hsa-miR-320a, hsa-miR-126-5p, hsa-miR-27a-3p, hsa-miR-30d- "Downregulation" selected from the group consisting of 5p, hsa-miR-10a-5p, hsa-miR-10b-5p, hsa-miR-497-5p, hsa-miR-134, and hsa-miR-150-5p. A method that is one of the "rated" miRNAs. The method of claim 1, wherein the upregulation of the "upregulated" miRNA compared to the control is an indicator of the subject having gastric cancer. The method of claim 2, wherein the down-regulation of the "down-regulated" miRNA compared to the control is an indicator of the subject having gastric cancer. A method of detecting or diagnosing a subject's gastric cancer stage, or a method of determining the likelihood that a subject is at the gastric cancer stage.
MicroRNAs (miRNAs) and at least 90 listed in at least one of the tables selected from the group consisting of Tables 8, 15, 16, and 17 in the non-cellular biofluid sample obtained from the subject. Includes the step of measuring the expression level of at least one miRNA having% sequence identity.
A method of diagnosing a subject as having one of stage 1, stage 2, stage 3, or stage 4 gastric cancer based on the differential expression of miRNA expression in the sample when compared to the control. The method according to any one of claims 1 to 4, wherein the control is a subject without gastric cancer. By up-regulation of the "up-regulated" miRNA compared to the control, the subject was diagnosed with stage 1 gastric cancer, and the "up-regulated" miRNA was hsa-miR-142-. 5p, hsa-miR-29c-3p, hsa-miR-29b-3p, hsa-miR-27a-5p, hsa-miR-101-3p, hsa-miR-590-5p, hsa-miR-21-3p, The method according to any one of claims 4 to 5, selected from the group consisting of and hsa-miR-106b-3p. By down-regulation of the "down-regulated" miRNA compared to the control, the subject was diagnosed with stage 1 gastric cancer, and the "down-regulated" miRNA was hsa-miR-103a-. 3p, hsa-miR-26a-5p, hsa-miR-146a-5p, hsa-miR-15b-5p, hsa-miR-362-5p, hsa-miR-671-3p, hsa-miR-454-3p, 4.-5 of claims 4-5, selected from the group consisting of hsa-miR-328, hsa-miR-221-3p, hsa-miR-23c, hsa-miR-136-5p, and hsa-miR-485-3p. The method according to any one item. By up-regulation of the "up-regulated" miRNA compared to the control, the subject was diagnosed with stage 2 gastric cancer, and the "up-regulated" miRNA was hsa-miR-142-. 5p, hsa-miR-29c-3p, hsa-miR-19a-3p, hsa-miR-93-5p, hsa-miR-140-5p, hsa-miR-148a-3p, hsa-miR-183-5p, hsa-miR-29b-3p, hsa-miR-424-5p, hsa-miR-27a-5p, hsa-miR-629-5p, hsa-miR-1280, hsa-miR-18b-5p, hsa-miR- 195-5p, hsa-miR-18a-3p, hsa-miR-550a-5p, hsa-miR-197-3p, hsa-miR-363-3p, hsa-miR-450a-5p, hsa-miR-20b- 5p, hsa-miR-15b-3p, hsa-miR-501-5p, hsa-miR-4306, hsa-miR-181a-2-3p, hsa-miR-16-5p, hsa-miR-128, hsa- miR-500a-3p, hsa-miR-501-3p, hsa-miR-629-3p, hsa-miR-25-3p, hsa-miR-29a-3p, hsa-miR-23a-5p, hsa-miR- 598, hsa-miR-186-5p, hsa-miR-93-3p, hsa-miR-23a-3p, hsa-miR-339-3p, hsa-miR-15a-5p, hsa-miR-140-3p, hsa-miR-29c-5p, hsa-miR-320b, hsa-miR-15b-5p, hsa-miR-221-3p, hsa-miR-29b-2-5p, hsa-miR-532-3p, hsa- The method according to any one of claims 4 to 5, selected from the group consisting of miR-374b-5p, hsa-miR-589-5p, and hsa-miR-106b-3p. By down-regulation of the "down-regulated" miRNA compared to the control, the subject was diagnosed with stage 2 gastric cancer, and the "down-regulated" miRNA was found in hsa-miR-107, hsa-miR-126-3p, hsa-miR-99b-5p, hsa-miR-339-5p, hsa-miR-320a, hsa-miR-337-5p, hsa-miR-193b-3p, and hsa-miR The method according to any one of claims 4 to 5, selected from the group consisting of −885-5p. By up-regulation of the "up-regulated" miRNA compared to the control, the subject was diagnosed with stage 3 gastric cancer and the "up-regulated" miRNA was found in hsa-miR-19a-3p, hsa-miR-140-5p, hsa-miR-148a-3p, hsa-miR-424-5p, hsa-miR-629-5p, hsa-miR-650, hsa-miR-1280, hsa-miR-27a- 5p, hsa-miR-18b-5p, hsa-miR-500a-3p, hsa-miR-629-3p, hsa-miR-550a-5p, hsa-miR-4306, hsa-miR-197-3p, hsa- miR-616-5p, hsa-miR-128, hsa-miR-450a-5p, hsa-miR-598, hsa-miR-15b-3p, hsa-miR-1290, hsa-miR-93-3p, hsa- miR-22-3p, hsa-miR-23a-5p, hsa-miR-320c, hsa-miR-130a-3p, hsa-miR-320b, hsa-miR-320e, hsa-miR-378a-3p, hsa- Consists of miR-9-5p, hsa-miR-29b-2-5p, hsa-miR-532-3p, hsa-miR-590-5p, hsa-miR-589-5p, and hsa-miR-29c-5p The method according to any one of claims 4 to 5, selected from the group. By down-regulation of the "down-regulated" miRNA compared to the control, the subject was diagnosed with stage 3 gastric cancer and the "down-regulated" miRNA was hsa-miR-30a-5p, Claims 4 to selected from the group consisting of hsa-miR-107, hsa-miR-126-3p, hsa-miR-99b-5p, hsa-miR-126-5p, and hsa-miR-27a-3p. The method according to any one of 5. By up-regulation of the "up-regulated" miRNA compared to the control, the subject was diagnosed with stage 4 gastric cancer, and the "up-regulated" miRNA was hsa-miR-142-5p. hsa-miR-29c-3p, hsa-miR-19a-3p, hsa-miR-93-5p, hsa-miR-140-5p, hsa-miR-148a-3p, hsa-miR-183-5p, hsa- miR-29b-3p, hsa-miR-424-5p, hsa-miR-27a-5p, hsa-miR-338-5p, hsa-miR-450a-5p, hsa-miR-579, hsa-miR-616- 5p, hsa-miR-200b-3p, hsa-miR-629-3p, hsa-miR-4306, hsa-miR-15b-3p, hsa-miR-627, hsa-miR-20b-5p, hsa-miR- 197-3p, hsa-miR-101-3p, hsa-miR-598, hsa-miR-23a-5p, hsa-miR-141-3p, hsa-miR-550a-5p, hsa-miR-18b-5p, hsa-miR-590-5p, hsa-miR-1280, hsa-miR-191-5p, hsa-miR-589-5p, hsa-miR-140-3p, hsa-miR-628-5p, hsa-miR- 93-3p, hsa-miR-106b-3p, hsa-miR-484, hsa-miR-29a-3p, hsa-miR-29b-2-5p, hsa-miR-29c-5p, hsa-miR-425- The method according to any one of claims 4 to 5, selected from the group consisting of 5p and hsa-miR-425-3p. By down-regulation of the "down-regulated" miRNA compared to the control, the subject was diagnosed with stage 4 gastric cancer and the "down-regulated" miRNA was hsa-miR-30a-5p, hsa-miR-107, hsa-miR-26a-5p, hsa-miR-126-3p, hsa-miR-99b-5p, hsa-miR-339-5p, hsa-miR-181a-5p, hsa-miR- 30d-5p, hsa-miR-30b-5p, hsa-miR-146a-5p, hsa-miR-10a-5p, hsa-miR-139-5p, hsa-miR-10b-5p, hsa-miR-23c, hsa-miR-497-5p, hsa-miR-154-5p, hsa-miR-382-5p, hsa-miR-134, hsa-miR-409-3p, hsa-miR-487b, hsa-miR-136- 5p, hsa-miR-150-5p, hsa-miR-193b-3p, hsa-miR-99a-5p, hsa-miR-337-5p, hsa-miR-495, hsa-miR-885-5p, and hsa The method according to any one of claims 4 to 5, selected from the group consisting of −miR-122-5p. The method is any one selected from the group consisting of miR-223-3p, miR-20a-5p, miR-17-5p, miR-106b-5p, miR-423-5p, and miR-21-5p. Or the method of claim 8, further comprising the step of comparing the expression levels of the plurality of miRNAs. 10. The method further comprises comparing the expression levels of any one or more miRNAs selected from the group consisting of miR-21-5pm, miR-223-3p, and miR-423-5p. The method described in. Expression of any one or more miRNAs selected from the group consisting of miR-21-5pm, miR-223-3pm, miR-20a-5pm, miR-106b-5p, and miR-17-5p. 12. The method of claim 12, further comprising a step of comparing levels. A method for detecting or diagnosing gastric cancer in a subject.
(A) A step of measuring the expression level of at least one miRNA in a non-cellular biofluid sample, wherein the miRNA is hsa-miR-103a-3p, hsa-miR-142-5p, hsa-miR-. 29c-3p, hsa-miR-30a-5p, hsa-miR-107, hsa-miR-26a-5p, hsa-miR-140-5p, hsa-miR-148a-3p, hsa-miR-29b-3p, hsa-miR-126-3p, hsa-miR-181a-5p, hsa-miR-27a-5p, hsa-miR-616-5p, hsa-miR-484, hsa-miR-4306, hsa-miR-590- 5p, hsa-miR-362-5p, hsa-miR-106b-3p, hsa-miR-497-5p, hsa-miR-18b-5p, hsa-miR-122-5p, hsa-miR-200b-3p, hsa-miR-197-3p, hsa-miR-486-5p, hsa-miR-99a-5p, hsa-miR-885-5p, hsa-miR-598, hsa-miR-454-3p, hsa-miR- 130a-3p, hsa-miR-150-5p, hsa-miR-30d-5p, hsa-miR-10b-5p, hsa-miR-532-3p, hsa-miR-23a-5p, hsa-miR-21- A step having at least 90% sequence identity with a miRNA selected from the group consisting of 3p, hsa-miR-136-5p, hsa-miR-1280, and hsa-miR-16-5p.
(B) A step of generating a score based on the expression level of the miRNA measured in step (a), and
(C) A step of predicting the likelihood that the subject has gastric cancer using the score.
Including the step of calculating the score by a classification algorithm comparing the expression level of the subject with that of a positive control (sample derived from a gastric cancer subject) or a negative control (sample derived from a subject without gastric cancer).
The score is given by the subject.
i. Have gastric cancer or have gastric cancer whose score falls within the positive control score
ii. A method for identifying the likelihood of having or not having gastric cancer (no gastric cancer), where the score falls within the negative control score. 17. The method of claim 17, further comprising measuring the expression level of at least one miRNA whose expression level has not changed in the subject when compared to the negative control. When compared with the negative control, the miRNA whose expression level has not changed in the subject is any one of the miRNAs listed in Table 20, hsa-miR-30e-5p, hsa-miR-. 340-5p, hsa-miR-532-5p, hsa-miR-23b-3p, hsa-miR-224-5p, hsa-miR-185-5p, hsa-miR-320d, hsa-miR-374a-5p, 15. The method of claim 18, selected from the group consisting of hsa-miR-584-5p and hsa-miR-194-5p. The method of any one of claims 17-19, wherein the classification algorithm is pretrained using the expression levels of the positive control and the negative control. Expression of any one or more miRNAs selected from the group consisting of miR-21-5p, miR-20a-5pm, miR-17-5p, miR-423-5p, and miR-223-3p. The method of any one of claims 17-20, further comprising the step of measuring the level. When the method is compared to the negative control, the expression level is unchanged in the subject and of any one or more miRNAs selected from the group consisting of miR-34a-5p and miR-27b-3p. 19. The method of claim 19, further comprising the step of measuring the expression level. The method according to any one of claims 1 to 22, wherein the subject is a Chinese.